ad-a234 792 - dticad-a234 792 g(l-tr-90-0317 satellite data derived estimates of erosion parameters...
TRANSCRIPT
AD-A234 792
G(L-TR-90-0317
SATELLITE DATA DERIVED ESTIMATES OF EROSION PARAMETERSFOR REENTRY VEHICLES
R. G. Isaacs APR 161991J. S. Belfiore
A ,9 1
H.-C. Huang
Atmospheric and Environmental Research, Inc.840 Memorial DriveCambridge, MA 02139
5 February 1991
Final ReportFebruary 1990-November 1990
Approved for public release; distribution unlimited
GEOPHYSICS LABORATORYAIR FORCE SYSTEMS COMMAND
DSTATES AIR FORCEHANSCOM AFB, MA 01731-5000
'1 4 15 136
"This technical report has been reviewed and is approved for publication"
GERALD W. FELDE JOSWH W/SNOW,Contract Manager Sa lute Meteorology Branch
FOR THE COMMANDER
SOBER A.McCLATCHEY, DirectorAtmospheric Sciences Division
This report has been reviewed by the ESD Public Affairs Office (PA) and isreleasable to the National Technical Information Service (NTIS).
Qualified requestors may obtain additional copies from the Defense TechnicalInformation Center. All others should apply to the National TechnicalInformation Center.
If your address has changed, or if you wish to be removed from the mailinglist, or if the addressee is no longer employed by your organization, pleasenotify GL/IMA, Hanscom AFB, MA 01731. This will assist us in maintaininga current mailing list.
Do not return copies of this report unless contractual obligations or noticeson a specific document requires that it be returned.
REPORT~ DOU ETTO A Eform ApprovedREPORT~~- DOU ETTINPGOMB No. 0704*O0188
t.roer-n '"Is cilctorr ,f mforr,'aticm s "stimafwO to avrg i hvur er resoorse including the timne for reviewing instrtr ions cr~~qe~tgOt hu~h. at nede Anl'~Q.fd <ofl'oietrrq ano rF-evnd the cI1ection of informationl sen commelh, rearrg t bre e mAt or ,,,n otherasueDOt Of ffh* e o*.mt. no ru go qestons tcr reoutno this burden I,, Vvahhrr'ton Heaoouartet ervices. Drectorate~o frrtO~e.r~vnfe~t.t1 fetr
s r... 1-1 -- q. *,.cn v-% .1;1014302 ad Jo th- fn.tof Mdn.,ement And Budget P~petork Reduction Project (O)O4-0188),WA..hington, Z'C 20503
1 AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3.REPORT TYPE AND DATES COVERED5 Februarky 1991 Final Report/Febrpary 1990-November 199C
4 TITLE AND SUBTITLE 5. FUNDING NUMBERS
Sdtellite Data Derived Estimates of Erosion Parameters for PE63311FReentry vehicles PR6670 TA17 WUB
_______________________________________________ Contract F19628-90-C-004E6. AUTHOR(S)
R.G. Isaacs, J.S. Belfiore, H.-C. Huang
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) B. PERFORMING ORGANIZATIONAtmospheric and Environmental Research, Inc. REPORT NUMBER840 Memorial DriveCambridge, MA 02139
9 SPONSORING.;MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/IMONITORINGAGENCY REPORT NUMBER
Geophysics Laboratory1Hanscom AFB, MA 01731-5000
GL-TR-90-0317Contract Manager: Gerald Felde/LYS
11. SUPPLEMENTARY NOTES
12a. DISTRIBUTION 'AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE
Approved for public release; distribution unlimited
13 ABST4,R V~ff V0tigates derivation of reentry vehicle erosion parameter estimates based] onthe inference of rain rate and hydrometeor liquid water content from microwave imagery data from theDefense Meteorological Satellite Program (DMSP) sensors. Specifically, this research focuses on theevaluation of Special Sensor Microwave/Imager (SSM/I) brightness temperature data to obtain aclimatology of intense precipitation at selected relevanit land based sites for a four month summerperiod. An integral element of the study is the,- identification of convective cloud models to support theanalysis of hydrometeor liquid water content from the inferred surface rainfall rates.
The study consisted of four tasks: (1) development of SSM/I rain rate climatology, (2)development of hydrometeor liquid water content parameterization, (3) application of liquid waterpararneterization, and (4) documentation.
Software was implemented to (a) read the SSM/1 TDRs from the acquired SSM/I data tapes, (b)calibrate and antenna pattern correct the data to obtain brightness temperatures, (c) bin the dataaccording to the location of each desired site by latitude and longitude, (d) apply the rain rate retrievalalgorithm to the data, and (e) evaluate relevant climatologies. The latter included four month summnertime series of average rainfall rate, spatial standard deviation, and hydrometeor integrated liquid watercontent for eleven Eurasian sites.15NUBROPAE
14. SUBJECT TERMS15NUBROPAE
Satellite Meteorology Precipitation SSM/I 161PIC2CD
Microwave Images Clouds DMSP 1.PIECD
17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20, LIMITATION OF ABSTRACTOF REPORT OF THIS PAGE OF ABSTRACT
Unclassified Unclassified Unclassifiea SARNJSN 7540-01-280-5500 Standard Form 298 (Rev 2-89)
Pretcribed by ANSf Std 139.t8298-102
cn
a,4o
D Disrib i-0
Dist?
TABLE OF CONTENTS
Page
1. Introduction .............................................................................. I
2. B ackground ............................................................................... 2
2.1 Rain Rate Retrieval iConcepts .................................................... 22.2 SSM/I Data Characteristics ...................................................... 32.3 Convective Cloud Models ....................................................... 5
3. Technical Approach ..................................................................... 6
3.1 Rain Rate Estimation Procedure ................................................. 63.2 SSMI Data Set .................................................................... 7
4. SSM/I Rain Rate Climatologies ........................................................ 8
4.1 Eurasian Sites ......................................... 84.2 Site Specific Rain Rate Climatologies .......................................... 84.3 Discussion ........................................................................ 20
5. Convective Cloud Model Parameterizations .......................................... 20
5.1 Cloud Process Overview ........................................................ 225.2 Review of Cloud Vertical Distribution Models ................................ 22
5.2.1 Simulation Studies ........................................................ 225.2.2 Measurement Studies ..................................................... 23
5.3 D iscussion ........................................................................ 245.4 Liquid Water Content/Rainrate Relationships ................................ 255.5 Summary .......................................................................... 28
6. Cloud Liquid Water ..................................................................... 29
7. Application to Selected Precipitation Events ......................................... 29
8. C onclusions .............................................................................. 38
9. Recommendations ....................................................................... 42
10. Acknowledgement ....................................................................... 44
11. R eferences ................................................................................ 44
Appendix A - SSM/I Data Extraction Software .................................... A-1Appendix B - Sample Output from the Tape Reading Algorithm ................ B-IAppendix C - Mapping Software .................................................... C-IAppendix D - Sample Mapping Software Output ............................ D-1Appendix E - SSM/I Data Catalog ..................................................... E-1
III
LIST OF FIGURES
Figure Page
1 SSM/I dual polarized 85.5 GHz brightness temperature vs. rain rate(mm/hr) with and without top layer of glaciated precipitation .................. 5
2 Functional flow diagram for SSM/I data erosion parameter estimationp rocedure .............................................................................. 6
3 AKTYUBINSK: Sample time series of SSM/I derived rain rate:(a) average rain rate, (b) standard deviation ....................................... 9
4 BLAGOVESCHENSK: Sample time series of SSM/I derived rain rate:(a) average rain rate, (b) standard deviation ....................................... 10
5 CHITA: Sample time series of SSM/I derived rain rate:(a) average rain rate, (b) standard deviation ....................................... I I
6 KIEV: Sample time series of SSM/I derived rain rate:(a) average rain rate, (b) standard deviation ....................................... 12
7 LENINGRAD: Sample time series of SSM/I derived rain rate:(a) average rain rate, (b) standard deviation ....................................... 13
8 MOSCOW: Sample time series of SSM/I derived rain rate:(a) average rain rate, (b) standard deviation ....................................... 14
9 MURMANSK: Sample time series of SSM/I derived rain rate:(a) average rain rate, (b) standard deviation ....................................... 15
10 PERM: Sample time series of SSM/I derived rain rate:(a) average rain rate, (b) standard deviation ....................................... 16
11 SEMIPALATINSK: Sample time series of SSM/I derived rain rate:(a) average rain rate, (b) standard deviation ....................................... 17
12 SIMFEROPOL: Sample time series of SSM/I derived rain rate:(a) average rain rate, (b) standard deviation ....................................... 18
13 TASHKENT: Sample time series of SSM/I derived rain rate:(a) average rain rate, (b) standard deviation ....................................... 19
14 Sample contour map of SSM/I derived rain rate ................................ 21
15 Simulation of the vertical distribution of the liquid water contentwith polynomial equations (see text) .............................................. 26
16 Cloud liquid water time scries for Tashkent ...................................... 30
17 Cloud tiquid water time series for Perm ............................................ 30
18 Cluster diagram for Aktyubinsk .................................................... 31
IV
LIST Ole FI(GURES
(continued)
Figure Page
19 Cluster diagram for Blagoveschensk .............................................. 31
20 Cluster diagram for Chita ........................................................... 32
21 Cluster diagram for Kiev ............................................................ 32
22 Cluster diagram for Leningrad ..................................................... 33
23 Cluster diagram for Moscow ....................................................... 33
24 Cluster diagram for Munnansk .................................................... 34
25 Cluster diagram for Penn ................................................. : ......... 34
26 Cluster diagram for Semipalatinsk .................................................
27 Cluster diagram for Simferopol .................................................... 35
28 Cluster diagram for Tashkent ...................................................... 36
29 Time series of ILWC for Moscow: (a) employing two region criteria,(b) assuming stratiform only, (c) assuming convective only .................... 37
30 Time series of ILWC for Blagoveschensk: (a) employing two regioncriteria, (b) assuming stratiform only, (c) assuming convective only .......... 39
31 Time series of ILWC for Leningrad: (a) employing two region criteria,(b) assuming stratiform only, (c) assuming convective only ................... 40
V
LISTI OF TABLES
Table Page
I Microwave Imiager Channel Applications ................................ 4
2 Candidate Convective Cloud Models .................................... 6
3 Polynomial Fit Data .................................................... 27
Vi
I INTROI)UCTIONReentry vehicle erosion is an adverse environmental impact on weapons delivery
systems with potential significance. This phenomenon which directly affects the accuracyof reentry trajectories is attributable to mechanical ablation of vehicle aerodynamic surfaces(luring reentry due to interaction with atmospheric ice clouds and solid or liquidhydrometeors (i.e. precipitation). Assessment of existing or prediction of future erosionseverity potential requires an understanding of the tempoitl and spatial distribution (i.e.climatology) of contributing meteorological conditions. These include both high altitudecirrus cloud and convective activity at lower altitudes. Ilus, appropriate analyses andobservation techniques are necessary to characterize cloud and hydrometeors. This reportfocuses on the latter issue. In particular, due to the site specific nature of the erosionproblem, analyses are required at specific land based locations.
Significant effort has been devoted to the development of analysis techniques basedon time/altitude cross section analyses to identify regions of likely convective activity(Feteris et al., 1976; Hardy, 1979). Iliese approaches supported by radiosonde andsurface station data were labor intensive and subject to the sparcity of available surface andupper air data. For the latter reason, for example, questions of spatial representativenessare always present (Bunting and Touart, 1980). Due to the level of effort involved inimplementing these approaches, they are unsuitable for application to large scaleclimatological studies.
Both cloud and hydrometeors can be observed globally using satellite based sensors(Isaacs et al., 1986). Previous application of visible and infrared satellite imagery data tothe cloud problem is discussed by Conover and Bunting (1977). The advent of remotelysensed microwave imagery such as that from the Defense Meteorological Satellite Program(DMSP) Special Sensor Microwave/Imager (SSM/1) provides the capability to directlysense hydroineteor liquid water content (Savage et al., 1987). Due to the radiative transferphysics relevant to the remote sensing of precipitation, the phenomenology of rain rateretrieval is different over land and oceans. This is due to the inherent differences in landtype and ocean surface emissivities (Isaacs et al., 1989). Retrieval techniques have beenidentified to provide quantitative measurements of rainfall rates over the ocean based onmicrowave imager data (Wilbeit c, al., 1977) and oceanic climatologies have beendeveloped (NASA, 1976).
Due to the spatial variation of emissivities over land and temporal features such assnow (which have signatures similar to precipitation), there are caveats related to theretrieval of rain rate over land. However, in optically thick situations when surfacecontributions to measured brightness temperatures are small, precipitation can be inferred.Such optically thick cases are those which characterize intense convection and these can bemapped over land (Spencer and Santek, 1985). The recently completed SSM/Icalibration/validation study provides further confidence in the validity of seekingquantitative inferences of rain rate over both ocean and land. The specific analysis toolavailable in this regard is the rate rate retrieval algorithm developed by the University ofWisconsin research group (Olson et al., 1988).
While rainfall rate brightness temperature relationships have been validated basedon the MJCS 154-86 requirements for precipitation measurements, the physics of thereentry erosion phenomenon indicates that a knowledge of the liquid water content profileis necessary. This can be achieved by adoption of a suitable model of the verticaldistribution of liquid water content resulting in the desired surface rainfall rate. Theselection of the meteorologically appropriate liquid water content vertical (listribution modelis importunt due to the relationship between liquid water content and radiometric signaturethrough the dropsize distribution (Falcone et al., 1979). For example, many investigatorscommonly use the Marshall Palmer (1948) relationship which was derived for stratifonnrain. Erosion severity is a function of precipitation severity and, therefore, convectivesituations are of greater potential interest. For these cases, other relationships are likelyvalid (Ulbrich, 1983; Willis and Tattleman, 1989).
This research study applies SSM/1 based precipitation r ;mote sensing technology tothe characterization of reentry vehicle erosion environments over selected sites. Threecomplementary areas of investigation were identified in our study proposal: (a) dataacquisition and application of the SSM/I rain rate retrieval algorithm to the development of afour month summer climatology of rain rate over selected sites with an emphasis on theidentification and analysis of heavy rain rate situations, (b) identification and assessment ofappropriate convective cloud models to establish a parameterization of the relationshipbetween surface rainfall rate and the vertical profle of hydrometeor liquid water content,and (c) testing of the parameterization of hydrometeor liquid water content on thedetermination of liquid water content rrofiles for selected ca es with emphasis on intenseconvection. These issues are explored in the following sect'.3rs.
2. BACKGROUND2.1 Rain Rate Retrieval Concepts
At microwave wavelengths, precipitation sized droplets provide a source ofatmospheric attenuation analogous to the effect of cloud droplets in the infrared spectrum(Savage, 1978; Falcone et al., 1979). This attenuation mechanism suggests a direct causalrelationship between rainfall and microwave atmospheric emission which has beenexploited to infer rainfall rate (Weinman and Wilheit, 1981). Furthermore, the presence ofprecipitation within the field of view of mie-owave sensors is itself of considerable interestsince the quality of resultant retrievals of other quantities is most certainly degraded (Liou etal., 1981). For example, microwave sounding brightness temperatures such as those fromthe SSM/T must be corrected for rain in much the same way as infrared radiances are forcloud. To provide a theoretical microwave brightness temperature/rainfall rate relationshipapplicable to rain rate retrieval, models of both the rain layer (including a rai layer heightand thickness) and of rain microphysics are required. These models determine the verticaldistribution of hydrometeor liquid water content. Both of these factors, of course, varysynoptically introducing some uncertainty into the retrieval process.
Collectively, these factors determine the atmosphere's scattering and absorptioncontribution to total satellite incident brightness temperature which is evaluated as a multiplescattering calculation assuming that the rain layer fills the field of view. Microwaveradiative transfer theory accounting for multiple scattering has been extensively applied tothe study of atmospheric precipitation and clouds (Weinman and Guetter, 1977; Wilheit etal., 1977; Tsang and Kong, 1977; Jin and Isaacs, 1985). The surface background againstwhich the rain is modeled consists of surface emission and surface reflected atmosphericcontributions, which must be included.
Over the ocean, surface emissivity is low and relatively uniform, providing goodcontrast for the quantification of precipitation. This approach was applied over the oceanby Wilheit et at. (1977) to data from the NIMBUS 5' Electronically Scanning MicrowaveSpectrometer (ESMR) operating at 19.35 GHz. Accuracies of 2 mm/h and a dynamicrange up to about 20 mm/h were achieved using ground based radar for validation. Resultsbased on retrievals obtained using the Seasat and Nimbus 7 SMMR instruments, forexample, indicate that the retrieved rainfall rates generally underestimate those measured atthe surface with rain gages (cf. Gloerson et at., 1984). Lipes (1982) found that the SeasatSMMR failed to detect showery precipitation associated with convective cloud inmidlatitudes and underestimated rain rates for heavy precipitation. This behavior wasattributed to both loss of incremental sensitivity to precipitation at higher rainfall rates andinsufficient sensor resolution (about 30 km at 37 GHz) in convective situations. Sensorresolution plays a role since intense precipitating cells with horizontal scales of a fewkilometers will not fill the microwave radiometer's field of view. Similar results wereobtained in midlatitudes by Alishouse (1983).
Theoretical calculations suggest that over land, higher frequency, dual polarizedmeasurements can distinguish atmospheric scattering due to rain from surface contributions(Savage and Weinman, 1975; Weinman and Guetter, 1977; Huang and Liou, 1983).
2
I lowever, in practice the dynamic range of measurable rainfall rates is so much reduced at37 GIz that these measurements are virtually useless and considerable ambiguity ofsurface and atmospheric contributions still exists. In addition to the problems of specifyingappropriate rain models and obtaining measurements over land, another problem arisesbecause the typical horizontal scale of precipitating cloud elements is generally much lessthan the field of view of the ESMR instrument (50 km). Integration over a large field ofview creates a negative bias which underestimates instantaneous rainfall rates. In spite ofthese difficulties, areas of intense precipitation can be identified (Spencer and Santek,1985). Based upon the above discussion, there will be some difficulties in providingaccurate rain rates in intense convective situations from microwave data alone. The rainfallrates in the higher ranges of the desired domain will almost always be due to energeticconvective cells with spatial scales on the order of a few kilometers which cannot be fillyresolved with the 25 km FOVs of the SSM/I instrument. In such cases it has been notedthat high rain rates are always associated with minima in the equivalent blackbodytemperature (EBBT) field observed from infrared imagers (Negri and Adler, 1987a,b).The converse of this is not true, however, i.e, low EBBTs can he due to other thanprecipitation, therefore, it is advantageous to exploit both microwave (for lower rain ratesand to identify the presence of intense convective precipitation) and infrared (at higher rainrates) data. At the higher rain rates (and to help in the determination of beam filling atlower rates), OLS infrared imager data can be used. The specific approach adopted couldbe based on that described by Adler and Negri, 1988. They used GOES imager data todelineate convective rain areas by searching for minima in the EBBT field and thenassigned rain rates based on the results from a one dimensional cloud model whichprovided the relationship between convective development (cloud top height) and theresulting rain rate. Stratiform rain rates were delineated based on EBBT threshold criteria.Due to level of effort constraints, however, this study will focus on the use of SSM/I dataalone to derive convective rainfall climatology.
2.2 SSM/I Data CharacteristicsThe Defense Meteorological Satellite Program (DMSP) Special Sensor
Microwave/Imager (SSM/I), launched on 19 June 1987, is a passive microwave radiometerwhich provides brightness temperature data of particular relevance to the monitoring ofglobal precipitation properties. The SSM/I sensor antenna observes seven microwavechannels (three dual polarized frequencies: 19.35, 37.0, and 85.5 GHz and a verticallypolarized channel at 22.235 GHz) and scans conically with an angle of incidence on thesurface of the earth of 53.1 degrees (Savage et al., 1987). Using a single atenna to spanthis frequency range results in a sensor field-of-view (FOV) which varies with frequency:about 50 km at the two lowest frequencies, 25 km at 37 GHz, and 13 km at 85.5 GHz.
The exact spatial resolution of each FOV depends on the definition used and theantenna response pattern. Based on the polar orbiting DMSP satellite, SSM/I datapotentially provides global coverage daily, with twice daily coverage possible at northernlatitudes due to orbital overlap. SSM/I data are archived through the DoD-NOAA SharedMetsat Processing agreement by NOAA/ NESDIS through the Cryospheric DataManagement System (CDMS) at the National Snow and Ice Data Center (see Weaver et al,1987). These data consist of both satellite data records (SDRs, i.e. the calibratedbrightness temperatures) and environmental 0 1 records (EDRs, i.e. the retrievedparameters). Table I illustrates the sensitivity SSM/I microwave imager channels to avariety of desired geophysical paraneters (the DMSP assigned priorities for theseparameters are in brackets).
The retrieval of precipitation from SSMv/i data follows the statistical method outlinedin Lo (1983) and used in simulation for SSM/ data sets in Jin and Isaacs (1987). Whilesimulation retrieval results suggest that the required accuracies of a few mm/h are possibleat the lower rain rates a number of important factors are often neglected. These include thedegree of beamfilling, the vertical extent of the precipitation, the rain drop size distribution,
3
Table 1. Microwave Inager Channel Applications
Sensor ParametersFrequency (GHz) 19.35 22.235 37 85.5Spatial resolution (kin) 50 50 25 12.5Sensitivity (K) 0.6 0.8 0.8 1.0
Geophysical Parameters,IPrioritylirecipitation (land) [5] oPrecipitation (ocean) [5] 0 x xSnow cover [0 j o xSea ice (extent, type) [1.8] 0 •Sea surface temperature [81 o o xWind speed (ocean) [41 o x x xAtmospheric water (total) [31 aoSoil Moisture [91 oVegetation (e.g. Aibedo [11 x o x xCloud Liquid Water [7] x x x
Channels which are important for determining each parameter are indicated using the following code:Critical; 0 = Important; x = Helpful.
the 'emperature profile through the precipitating layer, and the presence of glaciatedprecipitation which can significantly alter the brightness temperature signature. Thesefactors are often kept constant in simulations whereas they vary in the natural atmosphere.Uncertainties associated wid. these factors certainly impact the accuracy of rainfall ratedetermination, and it is perhaps more reasonable to expect that a few broad rainfall ratecategories are retrievable from the SSM/I. Retrievals of precipitation (both liquid andglaciated) and surface emissivity from simulated SSM/I data were discussed in a journalarticle by Jin and Isaacs (1987) which also described a specially developed multiplescattering model which was designed to simulate dual polarized brightness temperatures inthe presence of inhomogeneous, nonisothermal distributions of atmospheric precipitation.
The capability to simulate inhomogeneous (i.e, those varying with height)distributions of precipitation is necessary to treat the realistic variation of rain rate withaltitude within developing frontal systems and the phase change (from water to ice)occurring in convective situations. In that paper brightness temperature simulationsspecifically applicable to tie SSM/I were shown. Figure 1, for example, illustrates thedependence of simulated dual polarized 85.5 GHz brightness temperature on rain rate andthe presence of an upper layer of frozen precipitation. The figure inset illustrates a modelof the simulated atmosphere with a 5 km rain layer over the ocean surface mid either a 3 kmice layer or another 3 km rain layer above. It can be seen from these results that due toenhanced multiple scattering at this frequency, the brightness temperature is significantlylowered by the ice layer. At lower frequencies such as 19.35 GHz, the ice layer has littleor no effect and, therefore, using the multispectral SSM/I data, the phase of the upperlevels of precipitation can be identified in the retrieval/analysis procedure in addition to therain rate. Tis provides a method to probe the vertical structure of the precipitation.
A significant calibration/validation effort (Olson et al., 1988) has focused on theSSMiI precipitation retrieval algorithm. The resulting investigation provides simpleformulae to obtain rainfall rates over ocean and land and addresses the loss of the 85.5Vchannel data. The latter issue is of significance for the inference of rain rate over land. Wepropose to employ the most recent land regression coefficients obtained from theUniversity of Wisconsin group. The steps in the retrieval are essentially: (1) antennapat, m correct, (2) apply existing SSM/I precipitation screening logic, mid (3) apply
4
i i i I '
275
-D250
225W
with top ice loyer. 200
(n)Wz
175-
MeO 3km
o 0 O 5kmn00.000
OCEAN125
I I I I I2.5 5.0 7.5 10.0 12.5
RAIN RATE (mm/hr)
Figure 1. SSM/I dual polarized 85.5 0Hz brightness temperature vs. rain rate (mm/hr)with and without top layer of glaciated precipitation
regression formulae for rain rate. Step (1) is part of the coding available with thecompacted SSM/1 data from Remote Sensing System, Inc. (F. Wentz, personalcommunication). The coding for steps 2 and 3 will be obtained via GL from the Universityof Wisconsin. We note that due to the additional computational load imposed by theantenna pattern correction step, it might be prudent to develop a prescreening algorithm toidentify precipitating situations directly from the antenna temperatures themselves. If this ispossible, only those cases identified need be antenna pattern corrected.
2.3 Convective Cloud ModelsAs discussed in die previous section, the SSM/I retrieval algorithm provides rain
rate as an output parameter although the radiative transfer physics indicates that radiometerbrightness temperatures are fundamentally sensitive to hydrometer liquid water content(LWC). Assessment of reentry vehicle erosion indices require determination of ahydrometeor liquid water content vertical profile. The required interface between surfacerainfall rate and LWC profile is a model of the vertical distribution of hydrometeorsappropriate to the observed synoptic situation. Previous studies such as Peirce et al.( 975) have focused on the applicability of specific hydrometer LWC models for thispurpose. Notably, Falcone et al. (1979) have specified climatological cloud andhydrometeor liquid water content models applicable to the simulation of microwave andmillimeter wave data sets which provide strawman cloud model candidates as functions ofprecipitation intensity categories (i.e. light, moderate, heavy). Since the application drivenclimatological emphasis will be on the delineation of convective precipitation, our focusvill be on convective cloud models.
Table 2 provides a list of candidate models which characterize a range of synopticsituations, including: frontal rain, thunderstonms, tropical squall lines, and cumulustowers. For example, a simple one-dimensional convective cloud model is that based onthe work of Cotton (1972a,b) and Simpson and Wiggert (1969). The model dynamics are
5
Table 2. Candidate Convective Cloud Models
Synoptic Situation ReferenceThunderstorm Helmsfield and Fulton, 1988Frontal Rainband Rutledge and I lobbs, 1984Cumulus Tower Simpson and Wiggert, 1969Tropical Squall Line Tao and Simpson, 1989Tropical Stomi Wilheit et al., 1982Convective Cloud Cotton, 1972a,b
based on the conservation of vertical momentum including the buoyancy effects ofperturbation temperature and liquid water loading. Physical processes included in thissimple model are the latent heating by condensation, entrainment of environmental air,conversion of cloud liquid water to rain (and its fallout), and the freezing of condensate.The microphysical parameterizations are kept intentionally simple, both in the interests ofrun time and because of the likely uncertainties in the input data. In Section 5, we discussthe hydrometeor cloud model in greater detail and identify parameterization for stratifonnand convective precipitation.
3. TEClHNICAL APPROACH-3.1 Rain Rate Estimation Procedure
A functional flow diagram of the project analysis approach is shown in Figure 2.The essential steps are: (a) reading of the SSM/I TDRs from the acquired SSM/I datatapes, (b) calibration and antenna pattern corrections applied to the data to obtain brightness
SSM/I-A ReadData Tapes
APC Temperatures
GeographicalBinning
Rain RateRetrieval
Cloud Model
Figure 2. Functional flow diagramfror SSM/I data basederosion parameter estimation procedure
6
temperatures, (c) geographical binning of the data according to the location or each desiredsite by latitude and longitude, (d) application of the rain rate retrieval algorithm to the data,and (e) evaluation of relevant climatologies. The last step includes calculation of time seriesfor the four month period for the rain rate and integrated liquid water content ofprecipitation. The latter time series required adoption of appropriate hydrometeor cloudmodels which are described in Section 5.
The software for reading the SSM/I tapes was implemented on the Air ForceGeophysics Laboratory Interactive Meteorological System (AIMS) cluster (Gustafson andFelde, 1988). The unmodified software was run and its performance verified. Certainmodifications of the software were made to enhance this project's performance. The firstaspect of the software modification concerned the performance time of the entire package.In its unmodified form, the tape reading package took about 8 hours to read an entireSSM/I data tape. Since there are 32 tapes, at best a minimum of 32 tape reads will have tobe performed, which would translate into over 6 weeks of work just to extract the datanecessary for this project. After closely examining the individual modules andcommand/data flow design of the software, we optimized the performance time such that itnow take,,,, slightly more than 1 hour to read a data tape. The data extracted by the modifiedsoftware has been verified and validated compared to data extracted by the unmodifiedpackage. Software was developed that: (a) loaded the site locations into the program, and(b) filtered the input data stream from any SSM/I data tape such that it would determine"local" data points relevant for each site, and group them into respective data sets. The datais binned using the exact latitude/longitude coordinates of each site such that a regiondefined by a square, whose center is the point in question, is generated. Any data pointswhich fall within this defined region are considered associated with the point in question.Upon completion of this filtering, a file for each region is written, containing the dataobserved by the SSM/I for the time period in question. Software to read the SSM/I tapesand do the geographical binning by site is contained in Appendix A. A sample output fromthe tape reading algorithm is contained in Appendix B. Output are the time (number ofseconds since the SSM/I sensor was first turned on), latitude, longitude and brightnesstemperatures for 19.35 v,h, 22.235, and 37 v,h GHz. The 85.5 GHz channel is read in asecond pass due to the different sampling.
A data display software package was also prepared that utilizes NCAR and GKSroutines to project the user's choice of either an orthographic, or cylindrically equidistantprojection of the earth, and then superimposes either the data observed for a specific site, orthe surface scantrack for the SSM/[ sensor during the periods in question. This softwareis given in Appendix C. Sample map output illustrating the data density as a function ofsite at various scales is shown in Appendix D.
Software was also written to evaluate the desired climatologies. These elementsinclude: (a) the calculation of average rain rates and spatial standard deviations within eachsite region, and (b) the evaluation of liquid water content and integrated liquid water contenttime series based on the adopted hydrometeor cloud parameterization. This software is alsoprovided in Appendix A.
3.2 SSM/I Data SetA list of the SSM/I data used in this study is given in Appendix E. Provided are:
(a) the tape number, (b) the beginning and end time of data set, and (c) the number of files.The data Files were acquired as part of this study and are available for further
analysis. The software capabilities assembled and developed during the course of this pilotstudy, i.e. data reading, data categorization, data plotting, and data analysis, should begenerally applicable to the analysis of this SSM/I data set.
While no comprehensive examination of the data set was made for purposes otherthan those of this study, it should be noted that this global data set should be extremelyuseful for a variety of other study purposes.
7
4. SSM/I RAIN RATE CLIMATOLOGIES4.1 Eurasian Sites
The study focused on eleven USSR stations of the Environmental DefinitionProgram. Data on these sites was obtained from GL. The eleven sites are: (a)Aktyubinsk, (b) Blagoveschensk, (c) Chita, (d) Kiev, (e) Leningrad, (f) Mo- "ow, (g)Murmansk, (h) Perm, (i) Semipalatinsk, ,j) Simferopol, and (k) Tashkent. Tne latitudeand longitude coordinates for these sites used in the study can be found in subroutine"estreg" in Appendix A.
Of these sites, Murmansk is the northernmost and Tashkent is the southernmost.Blagoveschensk and Chita are in the far east bordering China. Leningrad, Moscow, andKiev are the most western sites. Murmansk is the only coastal site. For the purposes ofthis study all of the sites were treated as land based.
This study did not call for the collection or investigation of conventional datasources such as surface and upper air data which might be available for these locations orfor the comparison of the SSM/I derived rainfall rate climatologies with these data. Inretrospect, this is an important consideration. Essentially, the subsequent analysis assumesthat the statistical relationship between rain rate and brightness temperature inherent in theSSM/I algorithm regression coefficients (which were validated over the United States andthe United Kingdom), are equally valid over the set of Eurasian sites.
4.2 Site Specific Rain Rate ClimatologiesReferring to the data flow diagram presented in Figure 2, the next step in the data
processing is the coding and application of the University of Wisconsin SSM/I rain rateretrieval algorithm to the four month data set. This has been accomplished. Data for eachof the eleven sites was binned according to lat/long and time tag and rain rate retrievalswere performed as a function of field-of-view within 400 km boxes centered at each site.This bin size was selected somewhat arbitrarily, however, consideration was given tocapturing precipitation events of synoptic scale which passed in the vicinity of the site aswell as mesoscale/convective activity. Considerations of time-space sampling for area-averaged precipitation (WMO, 1985) were considered in formulating our approach,however, there was insufficient time to fully explore these issues.
In addition to the spectral information available in the SSM/I brightness temperaturedata used to derive the surface rainfall rates, it was recognized that the spatial distribution ofrain within the binned area could be used for the purpose of helping to characterize themeteorological properties of the situation. This information is particularly useful to aid inthe selection of an appropriate parameterization of precipitation liquid water content (bothvertical distribution and integrated) based on the SSM/I derived surface rain rates. It is theliquid water content which can be related to reentry vehicle erosion.
For this reason, both the average daily rain rate (defined as the simple arithmeticaverage of theindividual SSM/I derived rain rates falling within the site specific bin) andthe spatial standard deviation (SD) within the bin were evaluated to analyze the spatialcoherence of the rain rate field. These data were plotted as time series to produce rain rateclimatologies for each site for the four month period. Climatologies :or each site areillustrated in Figures 3-13. Illustrated are average rain rate (Figs 3a-I 3a) and standarddeviation (Figs. 3b-13b), respectively. The time series are labelled in days from 1 June1989, the beginning of the four month data set. All time series plots have been put on acommon scale (truncated at 2.5 mm/h) so that intercomparisons can be made.
Examining Figure 7a for Leningrad, it can be seen that there are obvious high (in arelative sense averaged over 400 km squares) rain rate situations (e.g. days 60-65) and lowrain rate days (e.g. 35-45). The standard deviations are also illustrative (Fig. 7b). Fordays 60-65, the high rain rates are accompanied by a large spatial standard deviation (alsodays 27-30, 75, and 82). This might be indicative of cellular convection. Moderate rainrates with smaller standard deviations might indicate more uniform precipitation. TheLeningrad data set does not show this behavior.
8
- AKTYUBINSK
-20 40 60 s0 100 120Number of 0Oys from 01-JUN-1989
Figure 3a. Sample time series ofSSM/I derived rain rate: average rain rate.
L
20 40 60 -80 100 120Number of Days from 01-JUN-i1989
Figure 3b. Sample time series of SSMII derived rain rate: standard deviation.
]9
BLAGO VESCHENSK2.5
2 15
2
30 .
01
CHITA
21.
2004 80 10012Number of Days from 01 -JUN-1989
Figure Sa. Sample tim series of SSMII derived rain rate: average ram rate.
Z...
20 40 60 80 100 120q Number of Days from 01 -JUN- 1989
Figure Sb. Sample time series of SSM/J derived rain rate: standard deviation.
* KiEfV2.5
=1,5
Z
0L
0.5
0
2 05
2 AL
00 20 40 60 so 100 120
Number of Days from 01 -JUN-i1989
Figre 6b. Sampl time series of SSM/I derived rain rate: stareiateo.
21
LENINGRAD
20 40 60 so 100 120Number of Days from 01-JUN-1989
Figure 7a. Sample time series of SSMII derived rain rate: average rain rate.
0 20 40 80 s0 100 120e Number of Days from 01-JUN-1989
Figure 7b. Sample time series of SSM/I derived rain rate: standard deviation.
13
2.5
2
0.5
Figre8a. Saue fe ers frM11drvdri ae average rain rae.
2.5
S 2
-15
-120
Number of Days from 01-JUN-1989
Figure 8b. Sample timie series of SSMII derived rain rate: standard deviation.
14
MURMANSK
-F
ii
*1~~~ Number of Days from 01 -JUN-1989 010
Figure 9a. Samle time series of SS Mu derived rain rate: average rain rate.
2
2005o - l -- n.,,,
20 40 60 80 100 120
Number of Days from 01-JUN-1989
Figure 9b. Sample time series of SSMI! derived rain rate: standard deviation.
15
PERM
25
2
z
5
.5
2
15
F) 0.5
0 20 40 60 so 100 120
Number of Days from 01-JUN-1989
Figure l0. Sample time series of SSMI derived rain rate: saergeiateo.
*~ 16
SEMIPALA TINSK
-20 40 80 s0 100 120Number of Days from 01 -JUN- 1989
Figure I a. Sample tim series of SSM/I derived rain rate: average raiF rate.
2 5
2
20 40 60 80 100 120
Number of Days from 01-JUN-1989
Figure Jib. Sample time series of SSM/J derived rain rate: standard deviation.
17
SIMFEROPOL25
V
Isa
0.5
*~ 0
0 •I I I I !
0 20 40 60 80 100 120
Number of Days from 01 -JUN-1989
F~gure 12a. Sample time series of SSM/I derived rain rate: average rain rate.
2.5
2
I I
g0,
0 I II !I -
0 20 40 80 80 200 120
Number of Days from 01-JUN-1989
Figure 12b. Sample time series of SSMIJ derived rain rare : standard deviation.
18
TASHKENT
020
0 0 40 60 s0 100 120Number of Days from 01 -JUN- 1989
Figure 13a. Sample time series of SSM/I derived rain rate: average rain rate.
Z
0. 20 40 60 80 100 120
Number of Days from 01 -JUN- 1989
Figure 13b. Sample time series of SSMII derived rain rate: standard deviation.
19
In addition to the time series, we can look at the contoured daily SSM/I derived rainrates for given sites. This gives the opportunity to study the structure of the precipitationfor a desired day at the resolution of the SSM/I footprints.. An example is shown in Figure14 for Moscow. Based on the time series data given in Figures 8a,b, the figure illustrates:(a) a low rain, high SD case (day 15, [4]), (b) a high rain, high SD case (day 44, 10.41),(c) a high rain, low SD case (day 61, 10.2]), and(d) a low rain, low SD case (day 95,10.11). The numbers in brackets are the contour intervals on each plot in mnflh. A heavyconvective cell can be seen in Figure 14a, but there is little precipitation elsewhere resultingin a very low average rate for the region. Cells of moderate intensity can be seen in Figures14b,c also. In these cases most of the region is active resulting in high average rainfallrates. In Figure 14d by comparison, there is light rain throughout the region (note thecontour interval changes). The importance of examining both the average rainfall rate andthe spatial standard deviation in characterizing an event is obvious.
4.3 DiscussionThe type of meteorological event resulting in precipitation is likely to be related to
the distribution of hydrometeor signatures within the region examined. High standarddeviation indicates that the precipitation is scattered over the area of the view box, e.g.convective type precipitation, and the low standard deviation means the rainfall is uniformover the area of the view box, e.g. stratiform type precipitation. If high rainfall rate isassociated with high standard deviation, it might be precipitation from individualthunderstorms. The passage of a frontal rainband, on the other hand, could becharacterized by high rainfall rate associated with low standard deviation. The associationof low rainfall rate with low standard deviation could mean the passage of a warm front,and/or precipitation from stratiform cloud. If low rainfall rate associates with high standarddeviation, it could mean that the weather system which causes precipitation passes throughonly part of the view box. The other possibilities of low rainfall rate with high standarddeviation could indicate that the weather system is too weak to produce a sufficient amountof precipitation and/or the environment is too dry, thus only part of the rainfall could bedetected at the surface when the weather system passes through the view box.
Regions of convective and stratiform precipitation are hard to define, becauseseveral well-developed and decaying convective cells might still overhang into thestratifomi region (Tao and Simpson, 1989). Therefore, it is hard to use the rain rate toclearly define the type of precipitation. From the data, we can see that the two stationslocated at the eastern part of the USSR have higher rainfall rate and higher standarddeviation. The difference between these two stations might be that individualthunderstorms occur in Blagoveschensk more than Chita. Although both data show thatthe frontal rainfall occurs very often at both sites. It might be that the Blagoveschensk siteis much closer to the ocean. Other sites have less rainfall rate and show the situation ofsteady precipitation occurs more frequently in these locations, despite the fact that someindividual thunderstorms (maybe afternoon thunderstorms) occurred in the period.
We employ these spatial coherence concepts to the application of our precipitationcloud models (Section 5) to the determination of integrated hydrometeor liquid watercontent in Section 7.
5. CONVECTIVE CLOUI) MODEL IPARAMElTIRZATIONSTask 2 required definition of cloud models to relate SSM/I derived surface rain rates
to precipitation liquid water content. We have completed this task by definingparameterizations of liquid water content vertical distribution and total integrated liquidwater content of precipitation for stratiform and convective rain with the capability to decidebetween tile two using the spatial coherence data derived from the SSM/I rain rate fields.
20
I IIM
2
5.1 Cloud Process OverviewThe fornation of raindrops or ice crystals involves complicated microphysical and
dynamical processes. The thermodynamical effects from the phase changes of water inturn affect the evolution of the weather system in which those processes are embedded.Several papers have studied the dependence of the formation of precipitation onmicrophysical and dynamical processes. The results show that both the microphysical andthe dynamical processes are important. For example, the growth and fallout of rain caninteract with the updraft, or can evaporate in tie subsaturated downdraft region, which inturn, affects formation of the hydrometeors. Since measurement of the microphysicalprocesses is difficult, the only way to simulate these processes is to parameterize them withrespect to the large scale dynamical motion. Due to recent theoretical developments andlaboratory experiments, several models have been proposed to treat these processes andsome of the results have been compared with the measurements from the field experiments.
In this study we conducted a survey of different hydrometeor cloud modelsfocusing on those which can be used to simulate the vertical distribution of liquid watercontent based on the surface precipitation rate. A simple parameterized relationship basedon climatological statistics is derived and the vertical distribution of the liquid water contentis expressed as a function of height and surface rainfall rate.
5.2 Review of Cloud Vertical Distribution Models
5.2.1 Simulation StudiesA high correlation between cloud top height and rainfall rate has been shown by
Zawadzki and Ro (1978) and Dennis et al. (1975). Adler and Mack (1984) used a onedimensional cloud model to study the relationship of the thunderstorm cloud height-rainfallrate and the cloud height-volume rainfall rate with satellite infrared data. The variation oftie vertical velocity and precipitation efficien, j has been shown to dominate both the slopesand the difference of the two curves which represent the relation between the cloud topheight and the rainfall rate. Information on the convection from numnerical model output isneeded to derive the cloud-rain relationship, and the vertical shear is considered to be theimportant parameter to estimate the volume rainfall rate,
Larger scale factors such as the synoptic envirenment, topographical situation, andthe location of the station with respect to tie thunderstorm or frontal rainband are importantin determining the desired relationship. Therefore, empirical rain estimation techniquesdeveloped in one area cannot be applied directly to other areas. Simple adjustments may beinadequate because of differences in slopes of rainfall rate-height in different'locations. Todetermine convective rain rate and the volume rain rate, tie moisture source, the verticalvelocity, and the rain efficiency are important. Additionally, the updraft area is important indetermining the volume rain rate. Ideally all of these data should be used to adjust the valueof the input parameters to get a more representative profile. Thus, full use of a numericalcloud model with the inclusion of the adequate dynamical processes is necessary.
Kessler (1959, 1961, 1963) devised paranieterized equations for mi crophysicalprocesses (cloud water and rain) with an assumed vertical motion profile and watergeneration function. Simpson and Wiggert (1969) used a one dimensional cloud modelwith the Kessler (1965) type of parameterization to study the precipitation in tropicalcumulus clouds. The microphysical processes include the autoconversion from cloudwater to precipitation water, collection and coalescence, terminal velocily, and fallout of theprecipitation with the evaporation due to entrainment of drier air in a downdraft. The initialconditions include the information from the sounding data: the saturation at cloud base orlifting condensation level at environment temperature, the excess of the temperature at cloudbase, and the vertical velocity at the cloud base. The assumptions made with the abovemodel rule out the feedback between the microphysical and dynamical processes in the
22
model. The new model corrects this problem and the results are more reasonable.I lowever, the treatment of the ice phase is insufficient and, therefore, the paranieterizationneeds to be updated in order to include the ice phase change effect. The results from thisstudy reveal that more complicated microphysical processes should be introduced and theinteraction with dynamical effects should be included. The amount of the rain whichreaches the ground as precipitation cannot be calculated in the context of this study and,therefore, we cannot use the information of the surface rainfall rate to derive the verticaldistribution of the liquid water content.
Cotton (I 972a, 1972b) studied precipitation processes within the supercooledcumuli environment, and the interaction between the microphysical process and clouddynamics. Precipitation formation in warm clouds (1972a) and a model which includes theice phase (1972b) have been studied. A more complicated parameterization to simulate themicrophysical processes in midlatitudes, which includes the processes of phase changebetween cloud ice, snow and graupel, cloud water and rain water has been developed byLin et al. (1983). This parameterization is used in the study by Rutledge and Ilobbs toinvestigate the precipitation within warm clouds (1983) and seeded ice phase (1984)environments. Three different precipitation zones categorized according to the surfaceprecipitation rate are described in the study. Their results are in a good agreement withthose from field experiments.
Tao and Simpson (1989) use the Kessler type of microphysical (cloud water andrain water) and Lin et al. (1983) parameterizations (cloud water, rain, cloud ice, snow andgraupel) to study the structure of tropical squall-type convective lines. Two-dimensionalmodels and three dimensional models have been used. The role of the ice-phase and themesoscale ascent in middle and high-levels has also been investigated. The model outputof the ice runs have been compared with that of the ice-free runs. The results show that theice-phase microphysical processes are crucial for a realistic stratiform structure and itsprecipitation statistics. Also, their results show that the mesoscale ascent in the middlelevel was the main mechanism responsible for the extended region of the stratiformprecipitation at the rear of the squall line, which has been suggested by Rutledge (1986).
5.2.2 Measurement StudiesWei et al. (1989) use five different methods to estimate path-integrated (columnar)
cloud liquid water. The methods include one-channel (31.65 GHz) and two-channel (20.6GHz and 31.65 GHz) physical retrievals, the standard method of linear statistical inversionusing two channels, and two statistical methods that proceed from an initial determinationof several empirical regressions between measured and computed quantities. Withbrightness temperature data and/or the absorption coefficients (for oxygen, the water vaporand the cloud), the optical thickness of the clear air and the cloud water can be calculated.The calculation of cloud water content with microwave radiometer data is encouraging,however, their results lack comparison with independent observations, e.g. radar.
During the period of interest, there is no precipitation occurring in the study of Weiet al. (1989), thus they only estimate the cloud water content without estimating theprecipitation amount. The study needs sounding data (vertical distribution of pressure,temperature, humidity, and cloud water content), and knowledge of the absorptioncoefficients to resolve the answer. Also, a hypothetical liquid water profile from anarchival sounding needs to be specified in order to calculate the radiative transfer. Theseconditions require almost the same effort of running a more sophisticated dynamical cloudmodel. Heymsfield and Fulton (1988) studied the measurement of precipitation inthunderstorms from high altitude remote aircraft. hi their studies, the comparisons betweenthe microwave data at 92 GHz and 183 GHz with the data from radar, lidar or GOES IRdota reveal that using 92 GHz microwave data to detect the rainfall area is advantageous.The liquid water content (LWC) or ice water content (IWC) can be calculated based on the
23
convective area, IWC for the snow aggregates in the convective region and IWC innonprecipitating anvil region. The relation of the rainfall rate with the vertical liquid and/orice structure is unclear in this study.
Wilheit et al. (1982) used the 19.35 GHz and 183 GHz data from a microwaveradiometer to study the precipitation in a tropical storm. The tendency of the rainfall ratecan be matched by the tendency of the brightness temperature at 92 GHz, and the 183 GHzproviding some information on the vertical extent of the frozen hydrometeors. This is amore direct way to determine the vertical distribution of the liquid water, and it is worthfurther study.5.3 Discussion
Simpson and Wiggert (1969), Cotton (1972a,b), Rutledge and Hobbs (1983,1984) and Tao and Simpson (1989) used cloud models to simulate the spatial distributionof the liquid water distribution. Since the vertical distribution of liquid water depends onthe stage of evolution, the position away from the core of the updraft and the type of thesystem, the dynamical processes are as important as the microphysical process. Forvelland Ogura (1988) found that the addition of ice was responsible for the achievement ofmore realistic scale features in the convective region of the simulated squall line. Morecomplex microphysical process which included the ice-phase parameterization have beensimulated by Lin et al. (1983). Rutledge and Hobbs (1984) and Tao and Simpson (1989)show that the model output is in good agreement with the field experiment data. The two-dimensional cloud model of Tao and Simpson (1989) is capable of simulating stratiformprecipitation and the areal coverage of the stratiform region. Therefore, if a numericalapproach is desired, it is recommended to use this cloud model for the stratiformprecipitation cases.
The model of the Rutledge and Hobbs (1984) is good in the situation of frontal rainbands and convective precipitation. In order to obtain the vertical structure of the liquidwater from model output, several initial conditions are needed. Using this model requiresthe specification of additional sources of information. Since the liquid water field frommodel output is specified at each grid point, a simple relationship Lbetween the verticaldistribution of liquid water field and surface rain fail rate is unnecessary. If we have morespecific description of the liquid water field, the snow field and the cloud water field, thetotal liquid water content in a column can be obtained by integrating vertically. Also, therainfall rate is determined by the conditions of the environment, such as the relativehumidity, the wind shear and updraft. Therefore, it is hard to find a simple relationshipwhich will satisfy all of the dynamical conditions in determining the vertical distribution ofliquid water depending on the surface rainfall rate alone.
An alternative approach to derive a simple relation between the surface rainfall rateand the vertical distribution of liquid water content is using the climatological record. Fromthe result of Adler and Mack (1984), the height-rainfall rate relation is different from placeto place. For example, high rainfall rate for moderate and small storms occurs in thecoastal regimes, while fairly deep convection in Midwest thunderstorms produces onlymoderate rain rate (Adler and Mack, 1984). Therefore, our derivation of the relationshipsbetween the surface rainfall rate and the vertical distribution of the liquid water contentbased on the climatological data depend on the geographical distribution too.
In the study of Adler and Mac '1984), the statistical retrievals of columnar liquidwater and water vapor are found to be more accurate than physical retrievals. Therefore,we will use the climatological data to derive the simple relationship based on the verticaldistribution of the liquid water content from Falcone et al. (1979). The important featuresof these distributions are (see Figure 15):
(1) A maximum liquid water content is located between the cloud top and thesurface (Rutledge and Hobbs, 1983, 1984; Tao and Simpson, 1989). From the modeloutput, the height of the maximum liquid water content is strongly influenced by the
24
vertical vdlocity, therefore, the maximum height of the cumulus convection is normallyhigher than the stratiform precipitation, and
(2) The amount of the liquid water content is nearly uniform below the cloud base,or the liquid water content below the freezing level in the convective region is almostconstant with height (Tao and Simpson, 1989).
5.4 Liquid Water ContenRainrate RelationshipsBased on the results from the diagnostic studies, a statistical polynomial fit like the
one used in Wei et al. (1989) is used to simulate the climatological profiles for the fourprecipitation models given in Falcone et al. (1979). Models I-IV (pp. 46-47) correspond totwo stratiform (I, II) and two convective (III, IV) models, respectively. The followingconditions are used to derive the coefficients of the polynomial fits:
* the liquid water content reaching the surface, Ms, is derived from the rain rateretrieved from the SSM/I microwave,
* the maximum liquid water content is Mm and is at height Zm,
* the liquid water content at the cloud top (Zt) is zero,
" the first derivation of the polynomial at the height (Zm) equals zero, and
• the first derivative of the polynomial at the surface equals zero.
These values are read from Figures 14-17 in Falcone et al. (1979), see Table 3.
The relationship between the rainfall rate and the liquid water contentcan be found in Falcone et al. (1979) and Simpson and Wiggert (1969). We use therelationship of Falcone et al. (1979)
M = 0.07 * RR **0.83,where M is the liquid water content (g/m 3) and RR is the rainfall rate (mm/h), and apply itat the surface and use Ms to denote the liquid water content here.
A fourth order polynomial fit is used to simulate convective precipitation (modelsIII and IV). The result, as shown in Fig. 15a, shows the simulation of the verticaldistribution of the liquid water content with the polynomial equation and the input values ofZt, Zm, Mm, Ms corresponding to the four precipitation categories from Falcone et al.(1979) (Models I-IV, pp 46-47), given in Table 3. The cloud top of curve 2 is lower thanthat specified byFalcone et al. (1979) in their Fig. 15. If we look at the curves carefully,we can find that the slope is nearly the same above and below the maximum height. Insteady rain cases like stratiform precipitation, most of the cloud water and the precipitationis located in the lower troposphere. The maximum liquid water accumulation is near I to 2km, the height of the maximum liquid water field is lower. Thus, if the slopes above andbelow the maximum height are the same, the liquid water content will vanish at a heightbelow the specified cloud top. For this reason, only curves 3 and 4 are used to modelconvective rain.
A third ord,-r polynomial equation is used to simulate the stratiform precipitation.Three tests have bLen run with the equations satisfying four out of the five given
25
a 7
6.
km
2.
0.1 I I f
0. 1. 2. 3. 1. S. 6. 7.Liquid water content (M; g/m3)
b) 7. -T-r
H-I km
2.
0.
Liquid water content (M; 91m3)J
Figure 15. Simulation of the vertical distribution of the liquid water content wit/h
polynomial equations (see text).
26
Table 3. Polynomial Fit Data
RR Zt Zm MmCurve (m/h) (kin) (kin) (g/m3)
1 1.25 1.5 0.75 1.12 5.00 4.0 1.50 2.23 12.50 5.0 2.50 4.84 15.00 6.1 3.50 6.2
conditions. Test one satisfies the conditions 1, 3, 4, 5, test two satisfies the condition 1, 2,3, 4 and test three satisfies conditions 1, 2, 3, 5. The test two results (Figure 15b) are thebest fit for the simulation of stratiform precipitation.
The equation for the vertical distribution for the convective type of precipitation(e.g. Falcone models III and IV), is:
LWC = a * z4 + b * z3 + c * z2 + d,
where
a 2 = +1 [(3zZ2 - 2z, 3 - z3)M - (3zz,2 - 2z,3)mj,a ~ = z Sz2 -2Z, 4 Z 3 + Z,,3Z,4
-1 [j4- 2z 2 m Z,2z,2.4)M,]
5ZmZt2 - 2z m4z 13 + Zm
3 Zt4 [(4 -2 - 2zm,- (4Z 2z,
-2 [(zM 4 -4ZZ3 + 3z 4)m, + (4z, z, 3 - 3z,")mlC ~ -" Z4t 2-2 Z,~ 33 + Z' 2Zt 4R t
d~m
Curves 3 and 4 in Figure 15a provide the vertical distribution for convective situations.The equation for the stratiform type of precipitation (e.g. Falcone models I and II)
is:
LWC=e* z3 + f* z2 + g* z + h,
where
27
e = 2- 2 [(Z, Z. M,+z, (2z. -,)m.]S Z1 (Z" -
Zm ;Z(Zt - Z.)
_ -1= 2 (,Z, ) - z,' 3)m, + (3m2Zt - 2,;)MMI
ZmZt(Zt - Zm) -3ZZ+2z),+(mZ?-2,)fm
h=m,
Curves I and 2 in Figure 15b provide the vertical distribution for stratiform situations fordifferent input parameters.
The vertical integrated liquid water content is:
ILWC = a/5 * Zt 5 + b/4 * Zt4 + c/3 * Zt3 + e * Zt + Ms
for the convective precipitation, and
ILWC = e/4 * Zt4 + f/3 * Zt3 + g/2 * Zt2 + h * Zt + MS
for the stratiform precipitation.Some studies show that the height of the cloud top and the surface rainfall rate are
closely related (Dennis et al., 1975). Thus we can derive the statistics of the relationshipfor different sites, and rewrite the cloud top as a function of the surface rainfall rate toderived a more precise vertical distribution of the liquid water field.
5.5 SummaryThese studies reveal that the formation and evolution of hydrometers have close
relationships with both microphysical and the dynamical processes. The inclusion of theice-phase parameterization is important in simulating these processes.
Since the vertical distribution of the liquid and/or ice content is highly dependent onthe type of the precipitation, the evolution stage of the thunderstorm, and the distance awayfrom the core of the thunderstorm (in the area near the convection core or in the anvil area),a simple relationship between the surface rainfall rate and vertical liquid water distributionis quite site and time specific. A climatological record of the different sites should becollected and useq to derive the simulation equation. The instant observation afforded bythe satellite qan be used to adjust some input parameters, e.g. cloud top and surface rainfallrate, which allow the sinulation to be as close as possible to the real situation.
The specific simulated liquid water field can be obtained from the results of anumerical cloud model, however, this requires a significant amount of additional data. Forthis reason, we have adopted vertical profile models based on climatological modelsrequiring a minimum of input data. The relationships provide a parameterization of vertical
28
hydron~eteor liquid water content with surface rain rate and other input such as the height ofthe cloud anJ location and magnitude of the maximum. The latter are also providcd by theFalcone et al. (1979) model classes.
6. (21)LJ ) LIQUII) WATERRecognizing the difficulties associated with low correlations between SSM/I
brightness temperatures and cloud liquid water content over land (and snow), the SSM/Icalibration validation team recommended that retrievals of cloud liquid water not beattempted over land (Alishouse et al., 1988). In this study we have specifically undertakento explore the parameterization of hydrometeor liquid water content vertical distribution andintegrated amnount for the purpose of characterizing intense convective activity. In theabsence of other data sources (e.g. visible and infrared satellite data), some correlation canbe inferred between precipitation events and cloud presence. 'lo that extent, climatologicalcloud models can be applied in an attempt to characterize the cloud liquid water contentbased on SSM/I determination of the hydrometeor properties.
To this end we have adorpted the cloud models proposed by Falcone et al. (1979)which accompany his respective precipitation event vertical profiles. A relationshipbetween rainfall rate and cloud liquid water of the fonn:
M = 0.05 * RR,
is proposed to provide a climatological cloud liquid water aajunct to the hydrometeor liquidwater content discut'sed in the previous section. Following Falcone et al (1979), thevertical distribution of the cloud is assumed to have the same vertical dependence as therain, except that its magnitude is sca!.,d by the above relationship. Results for two sites,Tashkent and Perm are illustrated in F';gures 16 and 17.
7. APPIICATION TO SELECTED PRECIPITATION EVENTSIn order to apply the liquid water content parameterizations for convective and
stratifonn situations defined in the previous sections, criteria had to be de,:.tloped based onthe mean areal rainfall rates and spatial standard deviation data available in the time series.This was accomplished by examining the ensemble properties of these two parameters forthe four month study period using a simple clustering approach. The mean areal rainfallrates and spatial standard deviation for each site were plotted against one another. Resultsare given in Figures 18-28. These figures were plotted on the same scale so thatcomparisons can be made among the sites. Note that area averaging reduces local valuesconsiderably. Based on these cluster diagrams an approach was developed to differentiatebetween stratiform and convective precipitation regimes.
Notably there is a general similarity in the behavior of the cluster diagrams. Tworegions are definable from tie data. In the first region, there is a systemmatic increase ofareal standard deviation with average rain rate. For some sites this increase is gradual andnear linear (see Chita, Figure 20), while for other sites it is rapid and nonlinear (seeLeningrad, Figure 22). The first region occupies the lower left quadrant of the availabledata. The second region occurs at higher rain rates and standard deviations than the firstregion, occupying more of the upper right quadrant of the data. The break points definingthe transitions from region one to region two vary with site. Based on our examination ofdaily contours, we have defined region one to consist largely of stratifoni events andregion two to consist largely of convective events. There was not sufficient time in thestudy to provide detailed analyses to support this hypothesis. This should be exploredusing conventional data sources.
The implications of the categorization defined above based on examination of tiecluster diagrams, is to provide criteria based on the rain rates and standard deviations todefine membership in one of the clusters (i.e. either stratiform or convective) and then use
29
2
15
05
PHo0 20 40 60 80 t00 120
Number of Days from 0 1-JUN- 1989
Figure 16. Cloud liquid water time series for Tashkent.
2
0 20 40 60 80 100 120Number of Days from 01-JUN-1989
Figure 17. Cloud liquid water time series for Perm.
30
4i
3
2I
A1
0&0 40 Average Rate of Rainfall (rnm/hr)
Figure 18. Cluster diagram for Atyubinsk
4
A &
0 3
Fiue. Clse digamfrlgvsht
N
aA
CA
Average Rote of Rainfall (mm/lr)
Figure 19. Cluster diagram for Blagoveschensk
31
4
cA
3&
2 3Aveag Roeo anfl m /
FiueA lse darmo ht
cA
.0
2A
o0
-. m
2 4
3 32
4&
3
2
01
S1Average Rate of Rainfal (mm/hij
Figure 22. Cluster diagram for Leningrad
33
4
3
E
C
o 2
C
.0
*0
m 1
0 1 2 3 4
Average Rote of Rainfall (mm/hr)
Figure 24. Cluster diagram for Murmansk
>A
A3 Az
S2
SA A
0AA0
o. AA
A A
0 I01 2 3 IAverage Rote of Roinfoll (mm/hr)
Figure 25. Cluster aiagram for Perm
34
0 214
Average Rate of Rainfall (mm/hr)
Figure 26. Cluster diagram for Semipalatinsk
31.
1 2 34Average Rote of Rainfall (mrn/hr)
Figure 27. Cluster diagram for Simferopol
35
'4
E
.2 2
0A
EA
0.Ao A
00
0 1 2 3
Average Rote of Rainfall (mm/hr)
Figure 28. Cluster diagram for Tashkent
the appropriate liquid water content relationship developed in Section 5 to evaluate timeseries of this and related quantities for each site.
We have applied this approach to the data for Moscow displayed in Figure 23. TheMoscow cluster diagram shows a region of low standard deviation (lays (SD < -1.3) withincreasing rain rate (region 1) bounded by high SD days. Thehighest SD days are thosefor which the rain rates are the highest. We have defined this-latter cluster as region 2 inthe context of the previous discussion. Therefore, the criteria adopted for this site is thatdays for which the SD is less than 1.3 are defined as stratifonn and others are convective.The time series corresponding to application of this criterion for Moscow is shown inFigure 29a. For comparison, Figures 29b and 29c show the same time series when it isassumed that all precipitation events are stratiforn or convective, respectively.
It should be noted that this criteria applies only to the Moscow time series and that amore general criteria for other sites has not been developed. Our approach would be todevelop the generalized criteria based on nonnalized values using the mean average rainfallrate (i.e. the mean of the time series of areal averages) and mean standard deviation (i.e. themean of the time series of spatial standard deviations).
36
17000
- F16000 L ]
o- 15000
00
13000
2000
0 20 40 60 80 100 2C
Number of Doys from 01-JUN-1989
Figure 29a. Time series of ILWCfor Moscow employing two region criteria
16300
- 16200
C
C
16100
V
16000
159001 1 1 1
U 20 40 60 80 100
Number of Doys from 01-JUN-1989
Figure 29b. Time series ofILWCfor Moscow assuming stratiform only
37
13200
E
13100
0
13000
.j
C,
12900
12800, I0 20 40 60 80 100
Number of 0oys from 01-JUN-1989
Figure 29c. Time series of ILWC fJr Moscow assuming convective only
To illustrate application of this approach to two additional sites the cluster diagramswere examined and Blagoveschensk and Leningrad (Figures 19 and 22, respectively) wereselected. The cluster structure for Blagoveschensk is much more diffuse than for Moscow.There is a much greater variation in standard deviation which is qualitatively manifested in amuch less organized stratiform arch. In this case we have based the definition ofconvective events on high average rainfall rate (> 1.5 mm/ih) rather than on standarddeviation as was done in the Moscow case. The results of applying this criterion for thedetermination of ILWC for the Blagoveschensk time series data are shown in Figures 30a-30c.
The cluster data for Leningrad exhibits yet another form of behavior. Rader thanbeing diffuse as in the previous case, there is a very tight cluster of low rain rate/lowstandard deviation near the origin with some outlying points of higher rain rate andstandard deviation. We have defined the former class as stratiform events and the latter asconvective. The results of applying these criteria for the determination of ILWC for theLeningrad time series data are shown in Figures 31 a-31 c.
8. CONCIUS IONSThis study has established the feasibility of establishing climatologies of stratiform
and convective precipitation events for site specific regions of interests based on the use ofSSM/I microwave imager brightness temperature data. The application of SSM/Imicrowave imager brightness temperature data to the determination of precipitationclimatologies for eleven selected sites has been investigated. "Ilie SSM/ precipitationretrieval algorithm has been applied to tie detennination of surface rainfall rates within a4(X) km region surrounding each site and additionally algorithms have been developed toprovide areal averaged rain rates and spatial standard deviations. A novel application of thespatial standard deviations provides infornation on spatial inhomogeneities of the retrievedrain rates in addition to exploiting the spectral information content of the microwavebrightness temperature data.
38
6000
5 0000
v
o 000
C-,
E
560
5 2000
10001
0 20 40 60 s0 100Number of Days from 01-JUN-1989
Figure 30. Time series of IL WCfor Blagoveschensk mploying toreioonlytri
6000
2000
EE} 1800
C
0
16000
"1400
1200
1000 I I t0 20 40 60 80 100
Number of Doys from 01 -JUN-1989
Figure 30c. Time series of ILWC for Blagoveschensk assuming convective only
6000
E2) 5000
C
o 4000C.)
"o 3000
,
0 2000
1000
0 I I
0 20 40 60 so
N~umber of Doys from 01-JUN-1989
Figure 31a. Time series of ILWC for Leningrad employing two region criteria
40
6000
C '
5800Z
C5,00
0
5200
50000 20 40 60 80 100
Number oi Days from 01-JUN-1989
Figure 31b. Time series of ILWC for Leningrad assuming stratiform only
2000 -Convective ILWC for Leningrod
C7
E
1800
C
C0
S1600
1400
1200
100010 20 40 60 80 100 120
Number of 0 ays from 01-JUN-1989
Figure 31c. Time series of ILWC for Leningrad assuming convective only
41
Time series of the areal averaged rain rates and spatial standard deviations havebeen calculated for a four month summer period for each of the sites. lliis cl imatology of'precipitation events can be used to provide an indication of intense convective activitywhich is of interest in the determination of reentry vehicle erosion. In order to distinguishconvective precipitation events, the areal averaged rainfall rates and standard deviations foreach site have been displayed in cluster diagrams for the entire four month period. Basedon these cluster diagrams, a set of preliminary criteria has been discussed to distinguishbetween stratiform and convective situations. Paruneterizations of the vertical distributionsof hydrometeor liquid water content for both stratiform and convective situations have beendeveloped based on climatological precipitation cloud models. These paraneterizationswere developed after review of the relevant dynamical cloud model literature. Furthermorethese vertical profiles of hydrometeor liquid water content have been integrated to providefunctional relationships between SSM/I retrieved surface rain rates and total hydrometeorliquid water content. Time series of the total hydrometeor liquid water content have alsobeen provided for selected sites.
It is recognized that cloud liquid water content profiles are also desired to calculatereentry vehicle erosion parameters. While cloud liquid water content is not directlyretrievable over land based on the use of the SSM/I brightness temperature data, modelbased and climatological correlations between tile vertical distribution of hydrometeor liquidwater content and that of cloud liquid water content allows for a rudimentarycharacterization of cloud liquid water. A simple approach correlating hydrometeor andcloud liquid water vertical distributions has been explored based on climatological modelsavailable from the microwave attenuation literature. 'Tliese nodels assumne that the cloudliquid water content is proportional to the rainfall rate and that the shape of the verticaldistribution of cloud liquid water is similar to that of the precipitation. Using thisapproach, time series of total cloud liquid water content is also provided for selected sites.
9. RECOMMENI)ATIONSAs proposed this study has focused exclusively on the use of SSM/I microwave
imager brightness temperature data to develop rain rate climatologies. Precipitation eventsare indicative of meteorological situations which are inherently problematic with respect toreentry vehicle erosion. A novel aspect of the investigation is the application of spatialcoherence concepts to the characterization of precipitation regimes such as stratiform andconvective which determine the vertical distribution of hydrometeor liquid water content.Rain also acts as a surrogate for the presence of cloud which is not easily measured overland with microwave sensors.
A number of avenues of additional study are recommended based on the resultsreported here:
* More work should he done to exploit the identification of convective precipitationusing the spatial information content of the microwave data. The exanliIation of the spatialcoherence properties of each site for the study period should be augmented withconventional observations to help in stratifying the observed arches and clusters accordingto the nature of tie precipitation event. For exanple, surface and upper air data should beavailable from ETAC to be used in the analysis;
* As noted in Section 5 it is possible to identify numerical models of precipitationdynamics to suppoit parmuneterization of rain rate/liquid water content relationships. In thisstudy we resorted to climatological relationships for expediency. A fully interactivemesoscale model could be used as a vehicle for satellite data fusion and retrieval purposes;
* lie algorithms developed here can be readily applied to the examination of alternatesites and changes in the statistical computations such as the time period and spatial
42
averaging field. Additionally, although bandpass limited by the inherent instantaneousfields-of-view of the SSM/I data, tile spatial power spectrum of precipitation events couldhe determined. Tlhis would provide additional degrees of freedom beyond the simplemeasure provided by the spatial standard deviation;
The clustering of rain rate and spatial standard deviations can be extended to usesurface observations to further stratify the precipitation regime characterization. Forexample, frontal and various thunderstorn categories could be added along with theirrespective vertical distribution models. Available surface observations should also beemployed to verify and validate the SSM/! time series. This was not investigated here;
Data fusion with sensor data sources colocated aboard the DMSP spacecraft (OLS,SSM/T, SSMI'-2) should be investigated. Visible and infrared imagery provide a meansto specify the presence of cloud, cloud cover, cloud type, and cloud top height. Theevolution of multispectral imagery will provide the capability to identify high, middle, andlow cloud in conjunction with nephanalysis techniques (d'Entremont, 1986). The use ofmicrowave and millimeter wave sounder channels provides a means to sense the verticaldistribution of liquid water content for precipitation analogous to the "slicing" approach forclouds employed using infrared sounder channels. The SSMIf microwave temperaturesounder provides a signature in each channel which is proportional to the presence ofprecipitation elements at altitudes characteristic of its weighting function. Grasiewski andStaelin (1989) have suggested utilizing the impact of precipitation on microwave oxygenband (60GHz) and line (118GHz) absorption to obtain precipitation vertical structureinformation. An application of these concepts to the profiling of hydrometeor liquid waterfrom DMSP SSM/T data could be fruitful in this rcgard. This would give a directmeasurement which frees one from the climatological precipitation vertical distributionmodels employed in this study. The vertical distribution of cloud liquid water can then beobtained using the simple climatological based correlations suggested in this study;
* The SSM/T-2 moisture sounder is also sensitive to precipitation, however, due toits shorter wavelength response, it also provides a cloud liquid water signature (lsaacs andDeblonde, 1987). Although the number of available channels (five) is not as extensive asthat potentially available from an infrared sounder, the millimeter wave channels respond tocloud liquid water content which can be directly related to cloud vertical structure anderosion parameter indices. The problem of high, variable surface emissivity inherent in theuse of the SSM/I imager data is mitigated by the use of the SSM/T-2 sounder data. This isdue to tile vertical resolution properties of the sounding weighting functions whichselectively sense the middle and upper tropospheric cloud liquid water (for increasinglyabsorbing channels of the sounder);
* Finally, it is recommended that the use of data available from civilian satellitesincluding both the Tiros polar orbiters and the GOES geosynchronous platforms beinvestigated. Use of multispectral cloud imagery from the Tiros Advanced Very HighResolution Radiometer (AVHRR) can be used for multispectral nephanalysis. Htighresolution Infrared Sounder (HIRS) data can be usedfor applying the C02 slicing method.The GOES VAS (VISSR Atmospheric sounder) provides visible and infrared imagery andinfrared sounder data. GOES infrared data could be particularly useful. The specificapproach adopted could be modeled on that described by Adler and Negri, 1988). "llheyused GOES imager data to delineate convective rain areas by searching for minima in theequivalent black body brightness temperature (EBBT) field and then assigned rain ratesbased on the results from a one dimensional cloud model which provided the relationshipbetween convective development (cloud top height) and the resulting rain rate. Stratiformrain rates were delineated based on EBBT threshold criteria.
43
14). ACKNOWLEDGEMENT'We thank Dr. Ken Champion and G. Felde of the Geophysics Laboratory for their
assistance during this effort.
11. REFERENCES
Adler, R. F. and R. A. Mack, 1984: Thunderstorm cloud height-rainfall rate relations foruse with satellite rainfall estimation techniques. J. Climate AppI. Meteor. 23, 280-296.
Adler, R. F. and A. J. Negri, 1988: A Satellite Infrared'Techniqlue to Est imate TropicalConvective and Stratiformi Rainfall., A. Appi2. Met., 2, 30-5 1.
Alishouse, J.C., 1983: Total lprecipitable water and rainfall determnination from tileSEASAT scanning multichannel microwave radiometer, J. Geophys. Res., (0),1929- 1935.
Alishouse, J.C, S. Snyder, J. Vongsathomn, and R. Ferraro (1988): Water vapor andcloud water validation in DMSP s~cial sensor inicrowave/imager calibrationl/validation.J. Hollinger (ed.). Naval Research Laboratory (Draft).
Atlas, D. and 0. Thiele (eds) , 1981: Precipitation measurements from space, Workshopreport, Goddard Space Flight Center, Greenbelt, MD.
Bunting, J.T., and C.N. Touart, 1980: Horizontal scale variations in satellite estimates ofweather erosion parameters for reentry systemns. AFGL-TIR-80-02(09. ADA092447
Conover, 3.11., and J.T. Bunting, 1977: Estimates from satellites of weather erosionparameters for reentry systems. AFGL-T[R-77-0260. ADA053654
Cotton, W.R., 1972a: Numerical simulation of precipitation development in supercooledcumuli, 1. Monthly Weather Rev.. 100, 757-764.
Cotton, W.R., 1 972b: Numerical simulation of precipitation development in supercooledcumul i, 2. Monthly Weather Rev., 1[00, 764-784.
Dennis, A. S., A. Koscielski, D. E. Cain, J. H. Hirsch and P. L. Smith, 1975: Analysisof radar observations of a randomized cloud seeding experiment. J. AppI. Meteor. 14,897-908.
d'Entreniont, R.P., 1986: Low- and midlevel cloud analysis using nighttime multispeptralimagery. J. Chini. AppI. Met., 25, 1853-1869.
Falcone, U.., L. V. Ahreu, and E.P. Shettle, 1979: Atmospheric attenuation of illimneterand submihliineter waves: Models and computer code, AFGL-T'R-79-0253. INTIS ADA084485]
Feteris, lP.J., A.S. Lisa, and A.J. Bussey, 1976: Environmental (lelinitimi programi crosssectionalanualysis: Summary of data mid analysis techniiques. AP7CRL-TR-76-00)2.ADA021908
Forvell, R. G., aid Y. Ogurai, 1988: Numerical simulation of a inidlatitude squall l ine intwo dimensions. J. Atmos. Sci. 45, 3846-3879.
44
Gloerson, P., D.J. Cavalieri, A.T.C. Chang, T.T. Wilheit, W.J. Campbell, O.M.Johannessen, K.B. Katsaros, K.F. Kunzi, D.B. Ross, D. Staelin, E.P.L. Windsor,F.Y. Barath, P. Gudmandsen, . Langhm, and R.O. Ramseier, 1984: A summary ofresults from the first NIMBUS 7 SMMR observations, J. Geophys. Res., 89(D4),5335-5344.
Grasiewski, A.J. and D.H. Staelin (1989): Numerical modeling of passive microwave 02observations over precipitation. Submitted to Radio Science.
Gustafson, G.B. and G.W. Felde, 1988: Interactive satellite image processing applied tocloud detection. Preprints: 4th conference on Interactive Infoniation ProcessingSystems, AMS, Boston, MA.
Hardy, K.R., 1979: Meteorological data analysis in support of the ABRES program.AFGL-TR-79-0084. ADA070946
Heymsfield, G.M. and R. Fulton, 1988: Comparison of high-altitude remote airtraftmeasurements with the radar structure of an Oklahoma thunder-storm: Implications forprecipitation estimation from space. Mo. Wea. Rev., 116, 1157.
Huang, R., and K.N. Liou, 1983: Polarized microwave radiation transfer in precipitatingcloudy atmospheres: Applications to window frequencies, J. Geophys. Res., 8(C6),3885-3893.
Isaacs, R. G. and G. Deblonde, 1987: Millimeter wave moisture sounding: the effect ofclouds. Radio Science, 22, 3,367-377.
Isaacs, R.G., 1989: A unified retrieval methodology for the DMSP meteorologicalsensors, in RSRM '87: Advances in Remote Sensing Retrieval Methods, A. Deepak,It.E. Fleming, and J.S. Theon (eds.). A. Deepak Publishing, pp. 203-214.
Isaacs, R.G., Y.-Q. Jin, R.D. Worsham, G. Deblonde, and V.J. Falcone, Jr. (1989): TheRADTRAN microwave surface emission models, IEEE Transactions on Geoscience midRemote Sensing, 27, 4, pp.433-440.
Isaacs, R.G., R.N. Hoffinan, and L.D. Kaplan (1986): Satellite remote sensing ofmeteorological paramneters for global numerical weather prediction, Reviews ofGeophysics, 24, 4, pp. 701-743.
Jin, Y.Q. and Isaacs, R.G., 1985: Thermal radiative transfer for inhomogeneousnonisothermal atmospheric precipitation and statistical retrieval of related parameters. InRemote Sensing Instrumentation: Technology for Science and Applications.Proceedings of the 1985 International Geoscience and Remote Sensing Symposium(IGARSS'85), Amherst, MA, 7-9 October, 1985. Univ. Massachusetts, IEEE.
Jin, Y.Q. and R.G. Isaacs, 1987: Simulation and Statistical Retrieval for InhomogeneousNon-isothemial Atmospheric Precipitation. J. Quantitative Spectroscopy and RadiativeTransfer, 37, 5, 461-469.
Kessler, E., 1959: Kinematic relations between wind and precipitation distributions, 1. J.Meteor., .f, 630-637.
45
Kessler, E., 1961: Kinematic relations between wind and precipitation distributions, II. J.Meteor., 18, 510-525.
Kessler, E., 1963: Elementary theory of associations between atmospheric motions anddistributions of water content. Mon. Wea. Rev., 91, 13-27.
Kessler, E., 1965: Microphysical paraneters in relation to tropical cloud and precipitationdistributions and their modification. Geofisica International, 5, 79-88.
Kessler, E., 1969: On the distribution and continuity of water substance in atmosphericcirculation. Meteor. Monogr., 2, Amer. Meteor. Soc., 84 pp.
Lin, Y. -L., R. D. Farley and H. D. Orville, 1983: Bulk parameterization of the snow fieldin a cloud model. J. Climate ApplI. Meteor., 22, 1065-1092.
Liou, K.N., G.C. Aufderhaar, and P.T. Nipko, 1981: Some examples of the effects ofclouds and precipitation on the temperature profile retrieval for DMSP SSM/Fmicrowave sounders, J. App. Meteorol., 20, 821-825.
Lipes, R.G., 1982: Description of SEASAT radiometer status and results. J. Geophys.Res., 87(C5), 3385-3396.
Lo, R. C., 1983: A Comprehensive Description of the Mission Sensor Microwave Imager(SSM/I) Environmental Parameter Extraction Algorithm, NRL Memorandum Report5199, Naval Research Laboratory, 52 pp.
Marshall, J.S. and W.M. Palmer, 1948: The distribution of raindrops with size. J.Meteor., 5, 165166.
NASA, 1976: Satellite derive global oceanic rainfidl atlas, NASA SP-410.
Negri, A.J. and R.F. Adler, 1987a: Infrared and Visible Satellite Rain Estimation. Part 1:A Grid Cell Approach. J. Clim. and Appl. Met., 26,1565-1576.
Negri, A.J. and R.F. Adler, 1987b: Infrared and Visible Satellite Rain Estimation. Pat IH:A Cloud Definition Approach. J. Clim. and Appl. Opt., 6,1565-1576.
Olson, W.S., F.J. LaFontaine, W.L. Smith, R.T. Merrill, B.A. Roth, and T. Achtor,(1988): Recommended algorithms for the retrieval of rainfall rates at midlatitudes usingthe SSM/I (DMSP-8). In DMSP special sensor microwave/inagercalibration/validation. J. ilollinger (ed.) Naval Research Laboratory (Draft).
Peirce, R.M., R.W. Lenhard and B.D. Weiss, 1975: Comparison study of models used toprescribe hydrometeor water content values. Al-CRL-TR-75-0470. ADAO 19633
Rutledge, S. A. and P. V. Hobbs, 1983: The inesoscale and microscale structure andorganization of clouds and precipitation in inidlatitude cyclones. VIII: A miodel for the"seeder-feeder" process in wani-frontal rainbands. J. Atmos. Sci., 40., 1185-1206.'
Rutledge, S.A., and P.V. Hobbs, 1984: The mesoscale and inicroscale structure mid'organization of clouds and precipitation in mid-latitude cyclones. XII: A diagnosticmodeling study of precipitation development in narrower frontal rainbands. J. Atmos.Sci. 4., 2949-2972.
46
Rutledge, S. A., 1986: A diagnostic modeling study of the stratiforn region associatedwith a tropical squall line. J. Atmos. Sci., 4.3, 1356-1377.
Savage, R. C., 1981: The DMSP Microwave Imager (SSM/1): Unfinished Plans. InPrecipitation Measurements from Space, D. Atlas and 0. Theile, (eds). pp 246-252.
Savage, R.C. and J.A. and Weinman, 1975: Preliminary calculations of the upwellingradiance from rain clouds at 37.0 and 19.35 GHz. Bulletin of the AmericanMeteorological Society, 56, 1272-1274.
Savage, R. C., L. A. Burgess, and B. W. Shipley, 1987: The Defense MeteorologicalSatellite Progran (I)MSP) Microwave Imager (SSM/I) in Passive Microwave Observingfrom Environmental Satellites. J. C. Fisher (ed.). NOAA Technical Report NESDIS35. pp. 186-202.
Simpson, j., and V. Wiggert, 1969: Models of precipitating cumulus towers. MonthlyWeatherRev., 97, 471-489.
Simpson, J., R. H. Simpson, D. A. Andrews and M. A. Eaton, 1965: Experimentalcumulus dynamics. Reviews of Geophysics, 3,387-431.
Spencer, R.W. md D.A. Santek, 1985: Measuring the global distribution of intenseconvection over land with passive microwave radiometry, J. Cli. Met.,.24, 860.
Tao, W.-K., and J. Simpson, 1989: Modeling study of a tropical squall-type convectiveline. J. Atmos. Sci., 46, 177-202.
Tsang, L., and J.A. Kong. 1977: Theory for thermal microwave emission from a boundedmedium containing spherical scatterers. J. App.Phsics, 48, 3593-3599.
Ulbrich, C.W., 1983: Natural variations in the an,!ytical forn of the raindrop sizedistribution. J. Appl. Meteorol., 22, 17(4-1775.
Weaver, R., C. Morris, R. G. Barry, 1987: Passive Microwave Data for Snow and IceResearch: Planned Products from the DMSP SSM/I System, EOS, 29 September, p.769.
Wei. Leighton, and Rogers, 1989: A comparison of several radiometric methods of,ciucing path integrated cloud liquid water. JAOT, 6, 1001-1012.
Weinman, J.A., and P.J. Guetter, 1977: Detennination of rainfall distribution frommicrowave radiation measured by the NIMBUS 6 ESMR, J. App. Meteorol., 16, 437-442.
Weinman, J.A., and T.T. Wilheit, 19, F. A survey of passive microwave and hybridremote sensing of precipitation, in Precipitation Measurements from Space WorkshopReport, edited by 1). Atlas and 0.W. Thiele, GSFC, pp. D-173-D-175.
Wilheit, T.T., A.T.C. Chang. M.S. V. Rao. E.B. Rodgers, and J.S. Theon, 1977:Satellite technique for quantitatively mapping rainfall rates over the oceans, J. Ap.Meteorol., 16, 551-560.
Wilheit, T.T., A.T.C. Chang, M.S.V. Rao, E.B. Rodgers, R.A. Nieman, B.M. Krupp,A.S. Milman, J.S. Stratigos and H. Siddalingaialh, 1982: Microwave radiometric
47
observations near 19.35, 92, and 183 GHz of Piecipitation in Tropical Stonn Cora. J_.AppI. Met., 21, 1137.
Willis, PT., and P. Tattleman, 1989: Drop-size distributions associated with intenserainfall. J. Appl. Meteor., 28, 3-15.
WMO, 1985: Review of requirements for area averaged precipitation data, etc. WCP-100,WMO/TD-NO. 115, World Meteorological Organization.
Zawadzki, I. 1., and C. V. Ro, 1978: Correlations between maximum rate of precipitationand inesoscale parameters. J. Appl. Meteor., 17, 1327-1344.
48
Appendix A SSM/I Data Extraction Software
A-1
.4
0.
4P.
-4
44
09b
ix0
0
OU
'Mo:
~~(4co zm B 04A040LI..T [
Q4 x ,
0, 14 4 9m h0w. W- C c 4
0. 0.6 14 4
00 u 004 4 94 0 0 OC>
r- x 0 14w 4.
0 > W- Q f 4 ow F.E4 Zw 0; IX-o . :Can z4 4 V> 14 IA 04 V 0 farw0 04 H -01 0 4.D. D Z 9~
. 4 x 0 vO C: .4 .4.1 (- 0 Ld1.0 0 104 .4 3 C U100 U3 1. 0. 4- 4 04
InP 4 NU V.4 0 .140 0 14 0 0Z w C!- 0 00,.~
.3. 934 t 0 0 14- V 1.3 0
04 :-F : 31 CO4 m- .E04 tr H a
9.3444 V)0 v x C 0 H414=4 .4 1. 4 .4~ 0~ U U " F
04 3 m 4=f 140 99 w .3 a .p4. V.. "4 z. .4 .4X Q" 1. z4 t
H 0 r. N449 U OAU w w w 0 0a0 1 0 U. 0 Ln inr' tq 14 E. t- 0
H4CI 0 N6. 4 , I 4 04 0 D4 . w 444..opc 41 0a 4J V o 'E4 f'.. 0o 04 4". F4~00 * .441
0 14 u 4 00 :10 ,q -i 0-'r1 :D.341 C .4. H9 L.C4.
wP. HP; 0 b 4 ) 4 a " 4 00t 01 00 94" A *.14O .40 04 00 1. >. t-14 w4 .44. x H0
16.14 0.. 0044 C a v14 'A . 0 .1f 4 14 .4 t-1V L004 E z p1 4 '. -~ a w 4 0 0(4 V . . .~1 .1411
*14.1 00 V)1~ H *0i .44 0. v11d01 ... u 0o..o 4 I 0 .14Q0"I =0 w@1 0 0.4 , 3 wn6n4 w w w161
014 w14 C430>t 4 N. U14. H E. W4 -. 4F444 C.- *1414U0041 " 0 0Iw *..0* E. Ff F414. >I6404 CO DO0 0 104 xx xD4'U0-04 4 :D ~ 4 4% 4 40. .44 -N4 Z U W I 4I
E.10 4 Ul 14 P c 140 V0 c1 WNN4 4 V)WW a.3Aj.
('I0 O ~ ~ U Z . -4 4 N H- It U ' 0 ' 0 t Iz 014 014 0 1)3 V4 . * 0% 94 4 E.4-444 r.-1q14it
1411 16 0 Li 40 1U. w o n' 0 Nc 0 . c..01t6161 .44* * *4O0140 40 4 4 V) m41 v4 W1 4 -N144411 w4 I.IllI
,41 14 M 4 VH 1000 MV0W 0 M4 4. 4140W l0 14 04 N 1M V 049 4 6 --r r .0 M - VZ
16 141 4 -4 0 NN 0I 0 In I
0.41400' * C.4A02
0 'A4
%0 H
A (a
0P 1-4 E.
if z
0
'U
00 0
0 W W 493 ,1
4 0 0
H 0 z14 U 0aA:4fl1 H 0 4 ..-
4~~. N .v4144..4 0
0; 04. w 0
C4' H1 0 :'L0 AU 00 >COH 0U 0jut 0mU
04 0000. H14 -X14 C64 .4u u'
0 'oc 0 .Q w
04400 f4 0
MO v 14 A 0 4v r, LW. = '4 NW44 E40 1. U.4
- 0 a44.2. H r 000 U- MXXX0 0 0
II co 04
'a 64 4 C44.0 H ~ > - 4
.I~0 0
MO H~H
4Z1.' a00 vi
a~~~~ 000f.:144H 4. 141414 .4 -4 4
0 PU t44N1 A 00 > 0 ' 0 a, 0.141 '.4
V0 0 0 0 vi 00 41..0HZ ~ '. 0W. F.44- -4 000 IV Q 0) 0 ) c
Z1. f! 0HH0. 41 V41 a 0 'a 000 I 41 0..00z 0 ~ 0 0 0 CA U 00o 04- 1. o -0o .V) wa, 0 0 4 N.'04 9V) In 4 1414 0 0 "44o .4 U.go
c 00' 000000-. w > 0U 1
0 '1. MN H 00 ~ .'A4-3
404
0901.
4:
H
ca
w
E4 0
1.4
C4
z 41.2
en z u
00
.4A-4
0
0:t
0
0
0
~00
4.45h4M.3
1-4C45400
4.34
0 W
C4 x 4
04 4 z0
LIID.
W04 f 03
0 0 04 1.4Z .3 1Z 04
5 .3 543 w v
0 3 0509 .3z u0, Z 3M )1 1,3Z 0 ;o 0H
4o X'0r H
W 0 4 0 tn00 40 N wm
W V v; '3 V
ZVS 0 D54 1 .m ( 5404.4 a,0
) tj .4 -W zVf S
* 4 W5 jW f' 4 54.44 H
a , ocu z!O .454IscS3
*0 0 :4 0 W !:a I'S 11I E.0, ,f* ~ ~~~ 0a E. z~0J N 5 33
Z -. 0V 0 z o3 0o
4 (- (4.4-0Z54 1-4unHVS .440.NZO0H .4 (4 V) a
* H- wm00 0o 1HO 5.3, 11z 0 5430--H.,'. ~ N :1 4 3
co ~ ~ 4'.-5 0 0oo u5(444 ~ 4
* 03 (-.'44.4 4H (4A-5
0
(4a:
ix00H
.<:>
9.3cc A
1 ; 44
94 9
= 0[4.4 .14 D t w-
1.40M z -0 0-004 0.3 6 t . f
0 6 4 x4 - n*
Z 9 0 Z. r 4 0
E- 4 :9 A C 9 0.!I I1 1
913 0 H 444t2 z Nf00 N0 W ow4 .4 1 a i A N00 [a4 C 0 6 4 H . 0-
.- 4. 9 9 1 ). uE4 . P 1..
II 0H P 4 , w0 UZ 0 0 C4 1.
>0 *44 0 H. *. .4 -.
WI x o9) 1 c % 6 0 3.4 U4 MM 0M0 94 PV 9 4 *0 MMC .
4 94 9 .4 z ..mt .4 oV 4 PU---(1 49. w 0 CH40f o z- 4 4 4 4424=0 - , o...
9194 ( 31. J 96 U U 3 .,- ' 00,4 .. 1
K 14 w P1w 4 1i4m U4 N .0 -W a q91 u g. 9 0 1 t. 2; 4424 q NN 01W
V4 03 m 04 . . 0 .1 -90 > 4 .4 9 0 .3-- Ox w .
C3 04 z 0 4 ". X.1.0 NN .
Vh Z4 V W aC 0 04 f: U4 rca to 0 4 m4 14 UUKC . M 1o
'4 ~0 0 K 0 14 424-( '-
14 . 4 4 C4 A U kE4 1 4 H Nf 9 4 4 a . l K1 4.46 t4 C6-K n - q
140 0 0 0 U .4 247 0 014'K K .4 P4 14 4 0 x- -Kf04. .-.- 014
94 (4 W . E. Z U3 J. 4 140 U1 X444.14 ( 4 A9. 0 412 9 4. 0. 0. 4 14. z z w w 0 . NM C .
W 4 Uw u2uNH u to) A;14 " 0:.C4 u 0 0 an 10 . ut
* n 0 14 w 4 m 4 9If 14 1 M9 v 0 UK 4-0% .0 N .. 4 m Ls42rc cN0 N. N- N4 4 N N- N N N M M M 4 M 4 M*M m m m 4 v4N vI v'9'n n% 0L
90 4 1 K 4-cc 0400 0 00 0 00 0 9 4 1.4044*1-0-4 U0
0 942 4 44242 14 * -14 N. -6' 9 * .0 K 4.4 ~ 0 94
0
40
(70
0
00> >4
M ni(r-Niso .
0~c0:P
040
04~ 4 -4A
4E.
m4%0
o& .4
0404 00 4 4 0HI x- Z ZM0000000 0w
U0. - 4 g-u u vI*all 4. H (4 N Z 0 1- ga4- ~ In *.w ww o *Z. c
0 0 .0 4 :t0> > 0000>M0 9 0 0 4.3 N44444H H c
U. E. (4 I.ZzE t D0 $1 0 0- H EHwH 0 1 0 ZZ WC 4
0~ 4 0 00000 4 - -w -:w * t03 z4 w44 4 4 4 W6 wZ ( s ri or o1 . 4 V
z- z .. z 4 z z4 (-Z- 14 .'z.M z0 0 Z. E, E ~z z z H 000 00 0 '
- -: 0 Eo- E-!0 4 -0 OD H H H H 0. 0V 1 4 .4 (4
~w C4 0. 4 14 00 00 0~~~'0~~~ J4( 04 0E4 El P. 6. f.44;44..- H ( 'H' II H n m44 =-91. Y 1-. . f-. II (4 6(o 0H 00 00 Z;~ N.t .: 4NP 4mM 4 :z 1(4 4 - to-H -( *4 w **44. El4.4.P*4*4*4 114 61..6 0
f4. .-4 (4'. w4-' .. . .!. ..0 00 0 . . *1 3
E-. E' 0~ 0-~ ~ 0 0 *(4 044 60~0 0 0 0~ 0. (4 60. * .HHHH q (4 0 011-.n W-1. 0E Z1. 0 (00 .(4H~~10 N41. 4 4I- I *~~( N 64 . (
.3 o0 ww-towr r,-( (4- C4-( .4 4 *E r.C -ww 0w0 w0 h( hm0 0 (4 *(4 0 -. 0Ca0v O 00.00 00 00 0*00 000 00 00 0Hi0I0000 00 00 00 000N
(40-(4 C-. (40.( 0(4 -- (C. 4 c . -744444444(.4 - ~ ' ..- U
Z0
0.0 I.S4.X .44 I414 04 34
40 0 ,01 .U00 0 0 0
0 U 000000
0 0000* * L.000000
t, Z w N N N NV
414 1
.4 00 040 644,0
14- o~~oc~oaa a14ax 44 4414 IC uU1S1
:> - 00100v 4uz 3
N: N1 w H
t'N00000 :0. a:.....
0 U 01.V004
000 0'v o.:
.4.4 ozz 000000Ii n 000000u . 4P 41 11 11 100 0. Cw axammaza N 44-W0. N
II14 H0
-4 X .0 S. a:.3.3.3.4.3
-4 A - V uu uxU 0 0 I)xz 43
x A. H 0 >>0 14 . 4P..i 4 -q S v v u0 0IV . 1 F4W
14 CI Z4; ( (
-t 0 a m: w a %:94 -4 U. >1 a: .0 ooo4. a H D 0 1
u 94- 4 ' 000000-. 31 4 3 11 - HHHH0 00P: Sn 4 0 0%.v cc C6 000Z -01 -. 04 00HHH
-Li 940 0 .Z 4 t40 S 0n 000000NH .H 94 4 3: 0 H . Ii I%~ -- A4 I IU S I
-; OZ -* li 6. .- 4r 40 4; 00(0N W 4 6- t4 FH V) 0- 0 00 0
0 -HZ9 U000 0
.-44 p f44 a: 0 H N M Mr 0 4 -z .4 -
0Ua:- -00 a: 0 (44,; V W U A.m P.a:CU:-
Q ti 4 M- j n0, Z ~HH P 4 0 z u z
P40x4 wl14 90 O0 ~ w 0
N-H- E. 0- Hl
*~- -- to41 0-4 *- 14 - 0
0'~W4Sn..24.~4.44 u 0 a 0 Lmi4 H-. 0 *4 N - L14 0 i 0 0 0 0 064. ~ 4l i i og~i - 10 u ow .4 14 D 0 4 00 0
Li30'~0: *H I . ' Ha 1 1 a 0 a: U o 0 0 0H H , 0
1..440~a~ac~W.4o OOO, .J Li 0. 0. *
0 C 4 r14 c
N N -4 -441 -4 44 04 It 0 0 0.4" 40 0 0
a U 000014 Li 4. 0 0 .4 0 000000
14 4 4 U0.4Z ~ 0 0 0 0 U 00 000
S0 000U'.0 cioo
004coo
n HN
14 r. 3O 4an 0 m000
04 ta 000 z P4 0: 000o0: * 000
C) if 000
V.t 0 .-4 .-4
00
0... W -
oou 0 0001., 000
000
m 4 '0. 4 .0 00
L. 0
L. 00004-VW 000 0
.4 coo
>I4 14C.4a 0
-4 4
04 0.
4 qmOKr-0 " .4 4 0.4
000000 14 00. Ci L. 000
044
* 400
V) 000
4 uNa:i O. M ..4
A-9.
0
046
14
0H.4
a
04
z 0
W 0
04.14z
Z 01OH (
v! A" .
04 4 .400.x4 t
00Z14Z94 t
14 0 1.10.404 .3
E4 A V 1~4 j
a 04 c
z u .4 . . .
1.3 1. w C g\ 2 i ZN .4
H Z > 141 4E4C
; . .4 mH NV) 4 ZX 1.4*4 0
C4 ~ ~ ' aH aU)X0WWU4F
w Z U4-O04 0 i EH 014040 n u N14 OC ZHZ1% 3 w .it 0" 0.~N w0:N -H0 0
.4 0 0 OOU1. ."4! 0 It.1F4k HZ Z t4P;
u (Y U 1.3 I.3 1.41z.30 44 :4 0 U) - WW 11Q"H 4. 2""9 ".
* w ot 0 < A z r a00m10 .4 13 10)E 0 V 'aQ 0(' U4N Z 0 0 0 f.
00 * 1 N 0G4Ua2ti.0a
co 0 042'.4 0 *
* 64 -4Z44.4.HAN10
0
0'4
F.
x ul
Oi/
vol
aj
LU
Coo
C u
u o
401
I.. ...
0(0
0 0
-4.-.
U , 00 0 0o
02 '4 , .' o oo
.0 • 4 * ".'*•00* '
* ..0 ,
0 * 4 :1022 '0 u a C
2 * Uo 0 Co
4 N
UU *. tn N0'
c :3 : :,, • n
1A-11
N 4.0 00 U A-4J0 IQ. 0A o
W, . .a c v 04.I
0 '4. a0 4ctl 00' 1.c 140ax ( n*..4 .0 r i 1 jma- 0 0 O 0NNc OC zl
w0*4 a .4 ,1 cV* d"r 0 0 0. 0a ' o. 444 . 4 . N 4 f n u c
V* f & V9 0 .4. a'4at rrC
*'.0 >. tn. 24, 14 -AO 204.4AA 0 'vtnw04c t o v _4 0 0 , t) 0r 0ru y nt-4
-U 4 o 1... .0 .0 o0 .0 '
M0 0 m L4 c o1 - 1 ,' .t .UL
'a 002 2.
.4)01~A- 1.
0
P:
14
inH
(4
900H
010
4.0
W144
40;
*'0 0
'o n n I fN Na 1,0N N NI0.4- - - - -- -- - - - -
0 0f tN m v : K l l l : .1: 9g000001wwA@o - - -0: :-0(7 hC
0.4 lWV W%10 t 0. ;4 - -0 0,; ; ; ... 0r wrrcc r ,a ) 9rrC M 1rCCCIr c
10 '4 0
wI ct4 o i"m r1wr ,o N r or o( 4 m4 w 1t -wa c N % ,. 7 c m4 0% 0 Jm-Lnk o% ot o -aw%0 I r -r-r ,r .m. o. 100 0 O. 0 .oo0g 0-
.0 .000 C .00 0 00 ...0 "' , -
0 000 0 0 0000 00 000000
* . .A.1.2
0
(.2
00
OD
p~0
.4.4
0.) 000r
0n 04 (4(41. m.4 N o 1
0.44
00 0 0
0403 041.2A
040 001 04 t
z V)*00
II 0
E- W 00114 a
.0 000 0 0
.0 f' H0 w 0 ,0C W'4! 0.00 JJ N V o000
0; tw 0.m vowt 0 0 -- a z [
0 4' -44 .4"0-C 4 .
N4 00000 0 0 9- k,
(A11
0
01 .
,
1*.
00
P1o.aU
.3 w
0 0
.- 4 144
0.4 4
0. '.N at.00p 0
Sn' 04 .4N 40 04
Z 4- 1 " 142: W o:Sn pn Snz
Z ~ :;;U3 .0 0.4
04NH si 0 u - nvC
Snv .4 HWS
t O 4 ~ Sn%
H Sn ZO HH O SnU
H U40 EH() i
14 0v f HH 0 1;, p PM104% l. 0.4 1
Sn 0 va O-Ocuwz~ 04P1t C4 a S
14~~A 14SnP1n
0
4%9400
(71
0
0o
000
94 +
.E4
04 04C
.3 f4 L O ; 4w0'. p 0. H
00 H tl 4 u 4=4 4 .. z r' !w Qc
040 04 0% - - 1%W 9 m--4- =H-19 ..4 I..II 0; cc I.. .
.30 H4 14 W-'N4 -E- A. m~ p 0n W ; ,1 :zI g a4 t.o .40 .4 ,4
.4 .4~9 Z;~ --- 4. HO >~ 'n -C 0N N. 60 g g n 4UF p 0 -- "" "- .0 3F ;9
H *~~g .404- --- 4U - H*~ 0. p~4 '0.4 - - 4
m~ 19 E 0; -40-''. 0 ic 0 :H 0 1. 0.4 24 -. 004
z In, p (A z0 0 m.0H; :C-1 w~ 0 ..
t4O 4.4
.0. 0. o-a 0'4v' *
O.E H 0 -4 v -. 0 -C - H -Zoo taC 03-' P- >- 0 0 -:
- ~.4N O rr P ) 1:- '- l 0 + ' M OC C .0. m. E 4 P. F4. - 4. p 4 - 4 . C '
r4w . .3 EC e. in H- 14' N j - -4 -4 of i . 00.44 . x4 4 '-00 0 0-- - F I 04 '- ' ta N -I
E. g (Cn C C H 4. OZm') m NzM.. . H 11 PW W H * 4 94 14 Z r ~ o m W -. 0 4 N 13 .4.f4 -m14 4 i
0 C V 0 C H H H . . N 4 * - - -g w .4 .. "U O '.w H Z 0 - 0 4 - -I
* . 9 0 U C 0 H . " 0 4 - 0 O W H Z~0
* 0 aH C a z H 1-3 NO=3U0- 4H 0 > *0 . I" 4 H -0 E E*N ~~~ ~ ~ 1 '(. WN E4 .40 tn.4 o C . .4 0 0..
V~ H.4 HH 0N0 p(."0- 0 > w 04- ' HHC~~~C~ to-4 CO'~x4 V~C H0-4 C4 CW - ~ ~ .4 .
CO.u I"-. -C . H.. .. 0 0 C. 0 0 P-
* ~ ~ ~ ~ ~ ~ ~ ~ ~ - 5'*40 * C H O N ~ 0 .0 l.~4)-H
0
0
[010
0H
aol*0+
to+
goo
w 0
14C0
go00
.44xw
I I z4
+~ +
IN,
.4 -14 +4W.10+ 0.D 0 E+ 7.Hlzm 4 0+ 1,:W)t
H4 .. 0 0 H 00
0 .4 -H -H- z : .Ho 0 N oco00 0 0 cc 0 0 c aV
-+ a4 H H(40 Vw0 0
"NM ow .40 " vo r o- mv wr 0 0
00 - 00 00o 00 0 00
00 0 0 00000.40 0 - .4
(4 -A416
0t 0
.~( m 1.'-4X
C0 U0 cc C, 0
1~~~ 14141414C,
L4 H4 N N M -f
0 . -4 0 014 0000
114 '0 0000u> f0IO f 14 0
*0 V . 3iE4~ (4 N N 4 93.E4.4
a:0 c0 Z HH$.4 lW 4
.4 0l000 0c a
41 1 3 414
u c 4 %4 4114I
044 0 t
1 14 00
0.44 .3. NOO
00 '0 00009 4 H -"4" Z14040' 0000t
00. N N 4.4v
04 - - - - - - - - - - - - -
11 oo 41- 9 .
.0 P0> I 444 C''.4 000 4 140*3042 00 000*.
U4 14 0 ca 0 w -
0 0..0 0 4.444 0000 Mut wH.H. H.)0F
0 0t"
9j 4 -F0 000 c4 .IV44U .4 o4 NO1 O
z ~ ~ ~ ~ ~ ' 0000.1, 0 o o o
0 00no zi
> .N44
'0 -421400. 4A N17
* S
-4
* 60
m4 u 04 4 0n
3r 6 094 0U :
4.4 0
40
aaE
04UU.
fn i
(nd 0 0
OH 0 0
inb 4 0
A-1
0
4
0:
014
U
V)a
41n
0OL'S.Z..
z -H z z
vi V m9 93 4
0~ 4 z0(n H
w00.94 a Iau 1420 4 0 (
04 P C049.
L4 X4.4 t xa a ,1%w wz .4 OH w z
0 n H to CH00N Z NH 54,4?4W0 H0
b1 w H .04 H .3
0 0 0~O
co w- .4 E- tb
a Uoz cwum~
%,= ZHOM040
m ~ E 0cv zz L IH0
0 0 0 . U. t 11 0 iiZ Z 0. H a U g
V. . .2 0 0 ()U- - f 1 1 F e!
z Sc o ~z4 0 .4 mZZZ IVO 09 0= 0.0..20r-Z H H 5lC
* 0 E I0aC.Sc z o 0. : 4dz w "I Clzc N, z 0 w04
co 0 0 H Z~0 .0
*0 v U.u4 04
* N ~~Sc44A H19
0
4
A
ix0
0H
0%L
044.3
m 0444m0k
0 Hcg
4 4aa Za .
mma4.1 0 0as01:I lg. 4 nL ,1 4Nm- n1.
P:0 9ZOH
r.t r : :r r : :ru qfIi .m 441 P.
on N
04 = 4 4 v 40 C1 4Mi M4 Nm I
0.4 .0 . . ...
0 a 3c N 4 ff
00 . * H 1.. M 1 0..40.0 41. 4 S. 1'4 IC , . V; .4. .44..4 40 40 04 40
z4- *01.44 .4 .4 1. %0 .4 MI -0 n%0I.U . 03 f4o 1. .4*~ 0.1 ! NU .4 -W~ 4 M~
'I 4.4 A .44 xU4 Q 4NMA.4 .. $ to 0 4q M 0n 0 Ch N~ m -'~ AAAC W in 0
0 V Z4/4 0 4I. .0 .4.. 4
O* 20: 0X 14. .4 w -* M r- In3. r: ~l ol ;4 NM 4t44H04 . 4 C3 . - 4 A. .4 .4 4v 4e Nnv0kOl z *I U' 0 4 H 12 U . 'aV [4 f Ho 0 4 '404 ?A O M M M O )444 444
.44 E: H C :"C4 0 . 0 -. w o4.!* 03 -.4.3t 03 ta p .4 .4. m n. ....
.Z 0:4 . 0 4O A 9 X0- F % m 4) .)000000404000r;*1 0, ,1 H4. 0 X E *- e "in N0.4441vaoo 43v)44no.3 1% X .MU NA %10 .4 04r.0 0. At14 0 0 >4 HO.1f
H 4 H4 U4 t. 6H 14 ., 0U)N, m w wMMoU44 44i44w4&410
4N 4 H 0 V Vo v u . .4 N 0 m4 &4
0 4~ V 400 H 04 01 -4 w *; ; : ..4'404. .3 9 H3 w 44 Z P 0C- 4 A C. nv me %* V4 I- ca 4p a 4 1' . 4 x .. 44 1. ? . .0 'D m w A 10?44?44
V) to z. o, = 4 z 4 z 413 0 .4H t. NO.4)'A. .4N4?1u. " - W4 k% N4 0 P; HO.u 0 .4 .4.*1.2~~~~~~*~ .4 4 0 * H 4 44
m m 4.4% w 7c ym i D-w , mm r00 0 m v w r 40 m%~ 0 m4 v1. 4 wn r-)A N40 m0 .N?4flw
4 . 0 00.'... Z00.40 0.40.000000U0000000000000000000000
U * 14. 12 144 '.(.0 4. A~.4 A4 20.
0
4
0
0:
4 a
0
00
(A0
10'1
f .4 +4t
m t1 c f
4N m
co z o1 lH .02
o. tn tnt 1 l .
aza 43os- w 0. N4 $0 4 4
m4 m 4- fnmI14w.II W c% c 14 14, 14 u-1 0HN aI
E. 14 +- z P. 11 o n 4
1-, a. o -4 .0: ca- 0 oO - 0o
-4-4~v 04.1 .43' . .4 .2u 0. , tO 0 w 14 o ::_. fo.-W* C I1-~' 1 44 4 4 1
4 P N4+- oN 14+.-) w-Cz1114 12-- lo04 o.
Ul~fl'ffl~fl1L) 14 * :040
14 1-. -0:0--C 11 O C4 P >
r-1CO4(h0-4 N 141 .4a. -- I4. *.4 .4u .4. 42 C AM 04 4 10 4
14a1 44.4 - 01 11 .4-* .. 41.01 . 0 - I H~ H14 0-
14 H1 14-.4 0.4 C41441
po .2.2 m40.403= A, o H14o~ +4 w4 H z4 !Oil, 0Z4 Z41Hw0
(a W. E z4 z4)1 . ' c * co 0 t0aN 140:-C1 UZ
(7 a, c14 a,1441E V 4 2H 14Z.4 r441 (4 U0 0 . '
r:0. I :1 0. m4 *, >I20 14 14 t4- O... 0 * 00
.o.14..4oca.o-4w m 4 011 P7 014 2 . 14 14 . 4 .0.H H 0.-1 H 0 . 40- .114C 1441 n..4- ti .14 .4 ~ 4 . ++ 1 4 I4 4
f4~144) H4 H. o L4. t 1wv4 N 144 (14 +--0 01 4 c% - -
c; ; C 4 4; 14 14 1i 1 o : 4 Z4 *( mMH w o ' N14: H 4- C. g, :> -H-w .4 .440 w .0
1.- >, 64 . .2 4 * z14 4. *424 4 go411 00o ca140014 Z 0441 W .NI.I1~)0
NIN NNN 1to1 a.0 014 u W~ 1441 ZO r 044-I** 4N 0
w..I4..4. o c14 4 4 0 11111 14 c44H 14 14Z1 *I2--. *. .0 N.2 m.V4.4444o I loW Wlo o 4 14 lo .14 1414 .21441 141.2--..2.2 owo ow c t(hMa mMac oaoo o4-1 o42 H 40 0 4 .~.424-
.-.-. 4 4 4 121444 OZ O 1 04 f4.)I212IIIII ..A1 * 021.2 Z
a.
0
0
01 u
ra
IA I
00
*10
44214
+ 4 . c3%
01.4.
40 "t
14o. oN 4 z.
0m * .3. N n 1 .
00 .14 14 w 01
040 I!. p 14414H a0tam A14 p .1:4
ti4 1414' 0 -C1 3. o - va0c
II 4 n Mo- w 0 . W ; :oO a12 w- Z: cc-t 0t nc
11 4 .' v4 + aa
zO a4 Z O Z41..2.~" 0 ar .1 440
C4 o- m to.4O *t0 I ---1.2 + 0 1
o12 o i 0.3* a
W OZ. um+ o -o - -o a o
* . 0 14 -. 04 ' .!zaWOc -W aA (4O Z .M 0- 1400I
4 xvi 0 -0 aa M. .4 . #.4a 4 01 N .. 4 "#l4Z H.ZZO 14 .3 ( 4t
.3 m 0-.4. F. w4 g.48. o a0ZO *0 14 a%.4. 0M 7 V ;; Om +0 -- 4. 0,1, o 4 *4 - N.3u
+. +~ 0.4 0 .413 CZ)4.
W 4'42 1 , 0 in4 -4. .. M 1 1 14 0 O 0 4 > ;4 .3aco4 aa . . 0.o-. 000 11 .3 - -O . 0
*oo 94i o -C 0~ 3 .3 .3 .3-4.4 .I.a . 0- c240.
C . z 4 a401 -0 m .3 14lI. -to + co H0 c C.443 t -[-- * .. +.I-00 0.4.
0 * o m U m.'.1 044 0 co &1 ~ - -a ~ 4
'A *4 0 Z Z M* . - . a . 1 . 0 1 1
4 '' *4* .4014 4.*0* a 00.04N4>OzZ~ *4 -0- -41. * 141420+?414041M14 v *u * . a4-. t . 02.-404t.44'*3.
ow4 w (ho *0' wkn01404.4 .4. *N *0. 04 D)14 o4l4.4N.l.4 oi-tIAX4 13C 4 0 0 + 0 - .' 0 01414*.-4 ,+ * *
*0* 4'2.4+*o A -22 4...--11-w4x4..
014
l. 0 J w
0 In. N0 kWin O0 400000..J
I. P.NNN-
.; 4
.A.
tC4 0; 04 Z.0 0~
%n t 0 -W V v 0 v -w
co xxx 00000000CIO U*w L. 00000000
000O O0000o00z*U 00000000
* 4J11.4 N n M v. N Nv
000
.4 0 13 .
ol4. 0 4.NZ0 1% 4 1% P P. .4
A .0> 0 E- >49
0. 46H H X .*Otto 0000 'A0NN 4~
CLt N 00 C6 l vv -
040 00000000a0A' -r W .0l ' mm oom.-w4N NW N~a
.%. 0 4. 41 lfl4no oc 1 - 4 4 00 N"v "cc4012.2. 00000000 L ni
$1400 00000000 t-44000000000 t
I 4 00000000
00000000II 0
00009 wz00m
44 .3.24.20a4 4 40 AF00 00* 000 a.0 .lc* w4 0A 4 a uV
z 4f 000 4 OQH44.21. 0 4 M U H "
t4 000 '4. NNI-
P ~ ~ 0 110% l9 -44vi L .Ww U oN N14 .3NN O .4 H 10 0 a n n flD L f2W.-4lo oo~
(4E- H 0 -l0 0 0 0 0 0 0 0 0 0 0 0tO 0 40 0 0 0 0 0 0 0 0 0 0 0
X f 000H.- 0 6 'Do ' 000100000 000N 000q0000040070 n -A 4 0 0 00 0 0 'a.4ooc o U 0000000000000.
w~~ ~~ 00000000 00 00 00 000 0000000
4 0 0 0 0 00N V N N v N N
z0Z~- 1.4. 4co40 .. ~ 042
.4 0 -. 0 0.~.40 .Z 40A-234CE..
01
o:0
0 .
00
4
0,
0;w-
04U4
443
04-Ik~S.
ao
0. 00 0
a 0 0
I~gN
V41 V q r c0,4 U
U IA
00A-20
10
0
14
90.0
1...4L3
.:1
*z 0
~w z
14 N
W .3
X(44
1a :'.4 -C
04 1414 V
04 0 .3
0.4 o w00-
00. 0 W4.4
4t10 w w4 1
wi u4 W00. r
.40 0(4 N
13)Z .44a z u 00r
N4 Z 4 Z %4 4 N H .4
C3' 0ww1 X 4 . Id4H ~ ~ :)0 t 0 0 0 3QW Z. N >NZO" 1.. E
*l 01 00 0 E. to to f!zo 00040 0 wf 0;-; Z940v3. o 0 :jv*F V Z-4z %j D H .V4t". _: 0 0 .204 "0 C491 4 - -
* .p 0-00(r4a0 .4 MOz 0 a o1uu z z N : .4 -2:::22 >0 0 z .w 1.. 11 N 1 11 1 .1. 1N 100 0 0 0O.~oo fv (4 (4.4(C :
* C. 02: I2A-25
14 i
0%0-44
0
ccM 1
0 0
40W t
.3.4 0 U-
14 u00 14 0 I
64 0 -N04-. .
z ca. 9 ; n"
n4 9 ; 0 4 0t H.
o.. -C in u 0. [ 4a,. 9 , U~o a P
40 . 4 u P 1.. r-040 w a X C; t0
.. 4.4 9 1 4 .ii~t 41 0 0 4V.
00 1 11 4N
4 94 0 p -0 4- >V, .a ,a A 0 4: .4 U- .
124 94 .0 9 H U4 0.q04,3 a> V0 4Et. 0W 4 90-A
W u 'I .: 9 H N 0 4. W 1t4934 0 9 . f4 go4 a:: "1
I . 9 4 a .g : -i - 1-0 N4C4 9P. 0 14 1 3 r . HAt4 U.,: - 9 1%1 U-
9 N 9 1 >N4N C= 9 990 *N N: a: -10u. H 9fs 9 w- - ."Ifq t
914 . Vu0 % 4 H- : P .0-- 9 +m >~9a:0t14~s ; - 4 i - N9 0 .
94.- 9 1 1wH&;4 P.o 3=I9'-a 914 : n 1Q r U) 9 ) 0 04 0 U4 :-0 ':.3ft Hc H H :
Cw - 40 X E a: 14 1.1 U4 - H 4 (4U =v- I
z ta w VA 0 D4 n~44 '0: U o W0 r 4: 4 4 4144u U X.. 0- M M mgoH 94. 0 3 4 4 t D
0. 40 w) Ha H~a =: 9 . OHH HHMD 0 9, z -4 .4 NO W 000 P. a
Vow H V 0 0: -a nrnw .a. 0po 4.4 N W1 0 0 4 OZ u . .454 x )104le.4 k. 04 a
M 1HP It a4 W. H4Pr C3 . f 14 Z.4 in40- 1 4 V: U.zP Z 4 w -0..W 4 :on 3~ H w )>Hc94 14 H U H - 99. 0 1H 0:1 Hh.N1 914 4-*
IA 14 9 IA 9 9 1 9 9 14 1--N9 e U 1 0 0
M v 0w 1 49 : i n -00ANn v 4n . I- No 0 E4-4 m49 v* 91wC4 Nmvowr o.no m na9~ H9 99 -41 919 N4 B.in-* I4 VOOM O
Z ~ ~ ~ ~ ~ 0 0 0 coo 0. -00 coo4 00.i09900-4 1M.U949 i 1000
4 9.O.. 1 .9 0H1 90 E U4 ,*itn.A-26 a
N t
+ 1
*14
fa 90
.4,14to'
a>
H 400m
> £ . oUx x11(
-w *
.4.C . .4'ou
R 0
M w a o83 4,, M tw $- -j:Hi
C.l [4 1ct ci v0..-S19 H PHZ= Ox o00 6 O 0 M1. X. Z v H 4 "
:D 0. 4W 0+wW F F0!:s ft[ 4q0 A m 4 z v o Z 9 = 9 P H H
z 4 H r ' ; .5." H H 4+ 0 u cif 3 3 .
:L Z o' 5-430( 00000=0= X 4 -0 ic c1 .VE *E .V o ~ N * w oc 0uH H 'H1 H H E
'Li0 040 q'5' .4 N V 4 5 4'4. * ZIn * c *~.
* w , ~ ' * 14 I
o~~~ ~ ~ ~ *N 0 N- mS 'Li U- w 1.1bf oa n or oV 4 mro ,c o Nmvtor qeU Z - . 4li .
5.5 -* H 4 f .4 Li* A -2 7
p.3
F4
3c 144I0 4 0,4
4 a: HHHN4M
a 0 ,
0 00000
*0000
0 H 4 0 0 0
.3 00000
m4.3
04 z
0 00
00100
[a a 49.4 0 0000004~01 0 0v 0 0 0
z z0. .000000
Olin 000000
0.4 0 04.3.4.N
I.W12 u 00000
.2. 00 00
00..04 94 a.I
:0 00>>
z 14HHHHX.
0* th 0.0,7. ,
.0 000'1 a',WHO 0 inntLN.4 000 0 0 0IA4. N
JJ~ 0000 2
0.~1 C6 A. I( *a N NN
F. NOOIA N 00400443 ~ ~ F 0N14.- 00 041 000 N NOo w O~
.. z4 0 ai 0 0 0 0 00 0 .3.3 .44N N4 '?0 ?00 H W 0 0 0 0 0 1J z .412w2.14 .4 V4
24 0000000000 00
w 0 00000V: 00 t 0%00000o
0 ~~ ~ ~ Na H. 0N U)" o00 0
0 41
002F. 0 000 0000.0. 0 01z P to tr 1.0.2a:00 cc
H1.Jz H 0 14 4 0 4 I 4 1 08 0 0 3.4000000.L4kaU 0 0 0 0.3 00 40.2 00 00 00 .J43. 4 . 00:I
44 H f I 0 w ..
0 In- 14 04 1-00a 0 z ca
4120~~~ 140tL 00 0 1
-44 I 1 0 0 .1.N.N1..4 I L.1.1.WSA-28.4 1.
4,
0,
Ln m
>0u
c-.0 0bi 0
ofl chw o
W 0
0 0OH o0 0 F
94. .. 0
.(.
u 0
0 ~ . 0 a .040 >4 4 ,
(4 P
0:41 .3 14
F400 0
co0 Z 0 4
0~0~ A020
0
0
x 4Ic.
004h
.
0 .
o 0)W mi
0.4000z
SO. 0 .3Z04 UC.)
C44 043
(4 '4
g. ( a4 U. 4 3
0 4 OH 00 '0 0t =U)4r3 lz 9
) 04
m 0 ~ 00 (
U) )EV .4
ZU4 0 '94 ~ 144 C6U)z o In 1 0
t4... 00aZt0 U1 0- 0 6
* - 000 000Lau-0 00U1 onz 0,z
*4 ..J a~ 4 U)c
* E0 wwm ~ x 1'. 0 . 4 Im 0 H
*4 z 0 ~ IAa02k.4 z~ 94
0 0 00 I I0 ~ H~C3UN 0 ~ 0. C
* Z --- U.0.H 0A) 30
0
x 4
0
0
44 "a
V
0W. 4
0..
AA H14VA .4-
$4.0av 443A040
OH0
os :0 n a.
X 1 9A
OM 'y t) 0 pC.0 .00O 0 '0 0 isauV
O 400 0 '\.04
*~ u4 C '4 0
MO 04 0 0 CIO
x a0 C4 m 0100C0 v C *-..4 V
0 OC. 0 C 4. 0 0 s
* 10,10 11 00'40 u C
'. 4. .0 '44. -10. 0 0
014 n UA 2 0 = -' -ii 0 4 u .4 4 's C* If. 0 c 0-. A.
13 4 C 'C 'fl0 0 ' 0
0 C 014. m N C4 NC m do C% a 4'N m v
"4 '444 0 C'' '014A4431
0p
0
0b. 0
a 14M4
X4 00 Hc
W 9.
l of 0P N 04
040
Q 04 ixP: 9.
4. 9
a,1hQ401
4 00
40 0
14140 0 1
MMz 0MN4 0 P "0" 00 IC9-
0 W. U 0141 z 0 0w
00 00 a O :0 zM00 v41 A. 0c 0z .00 0 .3 0
01.U 14160. P. N04
0441 0W U4
I0 04 o4WO 00 9 ZW
N 0t 0
.4. 140 400 P 00*
*00 0P 0-.OU
W0 14 14 N
00 004 4 0 -04 .4H.4 UU 4 0 ow 0 4 -NW
z 01 oz4 H 0 0 V(A3 ZU 4 .4w H W
94.0 0 04 = a4 0140 0 zf. 0~ z w. 0 4094 411 NV A -WH 0
U) 0 o N-0.40 -0 Ot 0 U. to 0 043 ~ ~ ~ ~ . JH 0 4'4 10
W. 00 4 0 u ; N-4 0- -100 n. 00 01 4 U
D 00D 0 O.0U4 10I .o ;tZ 0440 N
NN 0 0 =. u 14 E.MM4 14 14J 100a40 0 H 0V
UD 0 1. O 14V 1Z44U 1"0 0 ) 00 H 1140. N~)['
14 A , 0 0 W4 f14~C 0 1)1U0 4
14 0 0 W0 41 2 144 z I .V -a. 0 10 04 0N 0 a10oo m *--U.400N 0
x~1 OU.4000 0 0 ,.O3 0 M4 H00o ~4N 1o: 4' V) 'a U WP 0 . 4 4 4 0UU'14 Z' .C00 IC 0 14 0 C 13 0414
0o 0 0 4 0: C6 z 011440 N. Nwo 0: 0 44.-O
C4 N 4 N1 0 U.1 4 I.
4 44' 0~ ' . 014 0 0 004A-324
6
0
w0
0
0
Q• U
% H
0;
0
00
0 0
0p.,
O, ,°
o -
00
E v
0 0
0~ Z
U U
*0.
0 .
00 -
1)A-33
0
0
14
14
0H
40
0 *
~in4.cc W 9
400
01 0
94f 4lu L fP
0.- F! 0.A.
00 H I. -
F- 44 4 .4 IC4
0%0 4 w H ui03 -SO N +. 4g~X %0- 64 090
.4~~~0 on4 4 .4 U 4141400 4 In1 ON.to
CON. 0 p 04 .1t- 4V WHC H , ;! 1
4 It H 14 m H.4 -* . 4 0 04I)4
A;0 .14 UI IC I
414 H 1. .-- -
F4 p: N04-> H4 0 .3 -- 00 4 th
0IN4 0 14 W41441 r
X0- H4 U I - H14--
43 to as 3r 1 C p 4 c- 4P 64 E- a~nI p .4C C
0UN -C 12 E.-- 0 ! f 4C 0 i.4 ~ C O H 14. 0 n4n.~ .I40* - .44 El 00- "
41 4 22. ;.i 13 OZ-;It,A HD 4 4 Z4 R - 0-4 0 0 U 7, 64 0[4 4 0 U~ Hp >U:4 N +V Z H O P 14 10 rH 1414 0 +
4 'An 4 4 H 1 4.*4 .12 -4 ICON4 0~. 4 ~44 .
* 444 14*4 4 f -.4 . . .0 4 4 H- n IC I
13 4 0 4 . -C 144r U z44 0
*~90 Ow 40 HH D 41 .
.4 Cf 4 IH-.4 ZO0104 04 APO
14 0M 404 H U M141W 04 -4 N U M0 0 00H 0 4 ME 144 14 .4 4 NM
0 40 141440 0 0 4 N4H 1 14 0 .0000440000 0 40 00 0 14 44 14--. 0 0 00
O 14 C 14 0004 0014.41 0 00
II *~. .4 CU .144..4~1O.4~U 1 414 WZW1 A-34 44
,
go0.4
o 0. .41I<01. Ul 0.1
E4 E.b A 9.
(-4 ta 9 i c(70,wtaU 0
00
V0 00W
3410 .3.0
.4..
'A w t b
0~~ ~ ~ 0:a:016 t
:.Q 00>it t4H0C
0 q' w
C0010-kq'
0.04 :4 0000 3004X 0 0
0404V 0001.
0-- '0 00nxPzF L U. 144H0FaI .4 00 4. .V00 )a
00 P.3'l
II 0
Ii D V 00 0n*~~c 0110 0.
~~z OUO W 14.
A-35
0
90
0H
>
0 -CAH
IP~Do,
1434
W .314~14
040
0 .- X4 az mN 3 #
E. 0 O44 F
0; .. 1.3 9044 wUI i
940 w wx t
%, -44uo
0 0 H
t- 04O
0 > oc0 H
U440 H
Ck Z E E 0 0341 4[
w ZO 0HO0 U0
m 0 0a ) .01. o r '3 n0 o4 N f4". 44 P; -4 V
0 o*H 4 4 f 1.4 ..I.4 .3 & . .z uz-X IoZ1.. 1.2 4U
W M V) 0 00 z 1<a 40 .
z 0 unaw"XV0IflN :3*0 0'
H0 0 . 0 XHO A. k3 a.av 04~ . E UH 40~
* N 1 3U..H N - 0
0 >4033~ 0A-36
0
14
ell
0:0..
00
hi 0
44 0
a, In '
04
~Ic1- 0
0OH
>- 0
414-441
r. *4* z 0 N
4.. * I.4 - -4 Ui 0
0404~V P. a40404~~~ * c *I t' .
0 4 C4 . to*I 0 .
E0. *to *0
* 4.22c
U3 H a:i .2
E. 0' 00
to4 r aI 00a E.E'H 04 > a.0 M
Z 4W I- 1- 4 0 .x F4f.
M z400. 4D 14 0 94=
6H~~~~~~h ON V 4 ZC a u -E4~ 0 -' 0 1
-. F 400 SuH N h 0 40
to tai 4 c40 0 S'. - v4-. 00a .0 >4 1 v 0 0 0a0 v. .. 0U 0 ) c
U~ :..14 00
0
0 O141 144
14~C 00 *zI
'OH 4 l4.-44A-37
0
o 1 :5
In H N
H4 V400
.4 an0
40 *I 0
1f44 N40 1
4.414WW
00 4 x.Ch14w0
I.h,0 0
4 H44
144 M4 0
Z [4 4 14 3.H 0
.40 94 1F4 1) z
0.4 zI 2 0
0404 H A 00 a 14
.444 0 a4 u '0
042 4H 141
04 9. ow 0 4 1414
'.4 *14 14 *N 0 4 41 4 .4.14 H P H~ 00
IQ P2 14 ' 1
'4 H 4 W3 .4.4I14 H O m4 1.21.2 C
r4 g4 IA 44 =F I
.4 a. 0 x3 :5 4 C4W .0 0 U 1
'4 1 . 3 t'.4 00 tun I.w I 4 014'.6 P. 4 A
EC 0 .8 004X
a J V
* 01 04,..
11 11.240'
.1 Z O14 > wc, 04 0.4Do - -W f2 4 94 48 .3 0
N 440331414 1., t.1 04MH
4 1 0 H H20' 04[ 4DC ( 0 01.D C12 4 1 -1 0,* 04 '1.1.12 14 a0 0 oH9 4z>
04 0 F4O 3.3 1.1
-~~~~ u4 14.2 0-04 Is v 0 00 '
*~~~ 0 4U1,2W ' .. 4 .Ifm- n og c% 0Y 41 4 N , n% *c ,0 1NV
o N o Z o 4 W1,4.4 " -4 "9 NN0NF14 '0 00 0 00 0 000 9;1J ON .22.
4 3 0.1 '~'h~H..4W~ ' 4 >
C H ' Xl,114*4* ' -. 'A-384
z
I,.
H
IA
:0
IA
0
x W
in
0,-
U H
go
~A-A
GoO
-400
000
0000
A-3
(4.t4
Si.0
0o '
ID'D
0,
20 Uto 0
Ot 0 0 00
H (40.
V) 14 .3
Ci0 0d o
(40~S A (4.
C47 " 4 v .. :
0 41 u 0
" Si 4 1 1. H. Hi* l pa 09 0 0e
.. 4= 0. C4
M.1. 4. >4 C4(4@. -i 4 0 "? C40 4Q oD.
04 '1 14 Si cc(72F4 Si 1.
9i4 000S
Si~e 5-4 ZU2i 1
N o,, ;414 .2.4to (4 A c14- ~ mi M to1.2
V) 2d.4 23 4
U ' (4H~ W4 00 Hi
54 , ur 14. (4 m, v* ~ : i.. SiPi1i4S N4 0. (40. [4
o N Si C.24H 3 H2 0.4 V)o :c: '
-~~~~ :>0S ( 1
* 0 4(4 10. E4 f4 12I .452
oi 14 .4 goi go pt. Coa
.4 0 il 4* - .4 0
Si0 t. 0i 1.34S 0
Ff 00. P. li2'.1S 7 . .3 'a co
Si 000 0 Si 1- .4 (4 *Si .. ' N"N*~~~" 4 *1 S 4*.
* 4. .a -040- (A4(40
04..
0:
0 ..
(.3
x S
Ch
0S
0, 00in051w
OH
0
A-41
40
0
z $4
0
L1W vi 0
00 *2: W 0
In4 Hb H 1:1
00
;00 HA
0.o a 0m
021* N 0
in H 0.
? !:.an >.
444(400
.4 faC4014 HV)u
(4-~ 00 4
W~ ((40 4 94 4
W~ .4 000IA3 u v4 N:
'-40. W 400 >4
p 40 w 4D'.4 4 0: 4. 0hV r
0.4 14 4 gzw1 0 141 wI 0
00 *1 4.XX0
0.0 Z (4 4.t44 . 0
0: W H Mg 0404..~g 4.4 UW0000, 'o !1,; (4.
04 *44 - 4*24. :z 4 1 iEd 4.4 0' Z04 z4 1Z~0 C4 44 w ( 4 W1
0- H0 W 4. n0Ic: c1
' 0 0' .4.4'0.O E 00
f, Z4 Z* C'4M 4 WZ V
1=24 514 . I 14.4.02 (4) 0 :WII 0: 4.
In * 0 0 U w 0
4. 14 .4 00 4 0 >0t 4V 004 Z44 4.4 00 0
N m 4 v4 tn. Hw f-o0: 4Nmv
*~ OWN4.00000020000 0 00 0 00 o z41
4 14 W4O * 14 IA 4444. W £>
W >4 0: >* 24 0.A-42
0
00
a
0I
04U
flw H
co
4. :
01L0
0.40
000
000.
A-4
0
P
mU
0
0
4fl Hn
00
u w0c 14 M
U4 ") t
0 H4 F4
0C C 00~
0.4 01 9J Wto~ F4 0 ix 0%
00 z1 C 41 0.00 00 a. E. C 44
.4 * 4 0o1>4 z1 3 .,
00 H H u t4
a0 .
CC 4 003Z0 M4 43
*~~ to0 0
UO W C .3 00o
* C A H. m 44
*> zAC. P14 0 -040 :r toe m .
C) 54 0D4ca * 4 0 Z 01 ZC
Z Z > 94 w 0o Hm ~ C4 44 0v 04 tn 00.n C
14 4 14 >l~ 4 .4 4- -.E04 4P*a
if HH W ZM g V. :.. .24 t..
*~~~~ V 0 4 ~* v 4W 0~ 0H-14- 044 0
*: Z *.>4 4C44 0. .4 00 U*0 tH 0.4 0~~.2 0, 0 a 4 4 1 41
* 0 0 -4H4 H ~4 .4 N-4 O C0000000000000000000 0 4 Cd~a 4 ~
* 0000 000000 000004~ .I >4C .0
* 1. 141 H 014~0 H0(2 ' a ~ . .-.~A-44.
0
0.0
0
90
01 4
41n4
NN
4..0 w
94 1-4
0 -Cz4
-0 0 j
v0: 0 .
z S : . C404 14g au t
ZI 01. ( v
z 04 0
- XU4 o g NfA .3 44,Z ! m.
"I OW v 1 H.4"E
M Z 4 4 4H H14 4 z Z
z 0 a E4 I..
060M 1 Z2 X04. E.a
0 4 0 0
04-0zz ZH M 4 4
mm 4 9 W = 4 4 0 N,00Z 0Z Z. N)2 .
00 x.41N0 0N 0 0
00 u 00 a 0
* C.. .4 '-A-45
0
0
P0:P:0
P1 4-4P
mgI
o - 0 0z 0
OH
, m000
a,- 0 H00
z In
t41 404 4
04 4 643
V4 @04 4c 4.342
00 *.*0 r 0 1Auz34-A r AH 0 00
*.4~u X0 41~
SO. 1; :3 0
0. :1 O4 0 .44 A
It 44 4 0 0 4 ttnr NM 2 41 4.0 m - 4 .13c- %
-A. -A- - 4 -A -4 -A4 .I4 4 M
413 %4 W 43 - 4.
>,.U 0 043 00 0.4 04 00
a a.*~ a c 0 9: :9:
r44 C440V.C3.344:V 40 ta w44 0 0*44..4 V)40 0 la00
C t.0 4 .4''.41 :)2 3 ,'J30 4 UUVV* 41 A* 001 m4 1 41 43~1 a 44)4 H
* 1 0. * 0..O.C00- ~000
* 000 4 0 4:
1.4 0 000D0 0V000 4 41 41N N* 43'0 0 H0 0H0 0H0,0 0 0H0 0.04 0 04z4 0
V) 4 S) 0~144J43.I'4
w 0.n 0
* 14 4 V16 v. P. 000000000
4.-.4 00 0000 00 - N 0 4N 1 (
-. 10000000000000000 U4
''.4-A-46
0
0
0
x w
144
0:0
apz 0
0:14
.4 0
-. 4g
Ws 0 E4
01 -
N 4 04
0: 0 0
0 4 O 0VW .4
"
0
:10
N. 0: 0 : 0 4 C.4 Na 0:
z V)I- V vZ 0: . 10 &qC:
O20E ::0 0
t. O 1O 144r 1
w 4141-4 t; 0 0:W --
0 - v.
0: 20= a 00 0 4 0:0 E
0. 0 0:00.4:=:- 0NO 14001::1 U 44 111 F
:. F. Z: .64 14 14==I 4=" -0 .E.~ WM04wz0:0: .4 00O 0 14W=(-04N C: 00 0'0: z -0.1 0 ua w40cz o
0 0>E 0 0 *00 ti1.::4 : 4
0 Z4:4.4oss-A-47.
1A10.4
90oil
:>
10 Z00
14H
00
u1 Lua1 4 3 & z000 0
w1 H 4 F zC 9
w .zz 0 A 40zz P
4.w 3HHZ 4 gw 0 0Cn HN .
.4U. 3Uz 4 E 4 9
*14 p1 Z 2.
u 0. g ww94
04 Qwe4 0 H n0z , z 1
0Z1 - 4ZO.IIIIIt 0. (.1 Ic 4g C)4- 'A41 4t E4 I.w
0.-4 9 (4 H H4 H 34WWI -
000 " ~ 0 .4 ( Z Z H [4= 5 4 04
040 V D 0( 10 XZ 4 44 4 N 4 4 go N4 N040 1, 0:.40 - H *4 () I0r:4.4. 41 H(. Z 0 z. In z *g =
E4E 4 zwC o 4 f 1 4 114(* " ,4 v '40 H H 49
UL0 E.4~ (40 : 9 0( 9 ,4 H V.4 * '@0 r. co( 00 Z H H0 0 Z 0Z 62 tnN
0~a Ch w o H z z u 4 9 H . = H 1 ' 0 4444
. ~4 w 1.,(4 9( ( p~ C 41 - H H
1 * (40. W 4 ( a(4 Z Z: H% a1 .4 00 94 (4Z H* - H-d 0:44 1-4 W24 14 z.( : :1* 444
-~~4 0 Z:4(~N . 0 (4 W -,- 0
w z 0~40 a.4:4-44 0 0( -4 0 MCI c- 0 0: 0
*i " 0.1 44.M 4414O: (0 H 0 z .: .N H H 4* 0H.( 00(w z: L4 z .MUVZ-------ZaI4 Z E.41O 4 Z
C. 41 .rt 4H 14 N NU3W 9 . 40 N w9I 0 ( 4 H (4 0%010 (.4 (4;
"x *H (4WW WW W 0.4(4 (400 : 4 O 0) H
W: * W -"L00 P wZ O N N N O -4 4H I4. "
Z4 4 N00.4 (H H44(4(44(. H O 4 .W g 1 N MMM 4 p- j
w* p4H1 H4 m.4 O0O H H r 4- 0W4(4 (4~ 41c N. 94 (4=
0~~~~~~~~ * ((0 '.000004(4 H (04 0 4 0 (o (4
HY 4 H(40 H 40 Z:0 M04: 41 0 44
w 1 -g a o (44. WW.C. 4 010 0 4 H 0:
Z. 41 W H 4 H((4 H 04H0 (4 (4Z -9 (4 Z 41 H N4'.
14 w0H
0 (44( -0 -(40o (4- 04- H 0. H ( 4 '4f440
D r 41 ) -' : 3 (H(( :c~0.U 0.4 (4. OH, :I H 0-4-4 4
0 *-((2.H w(04H4(00 0:04 0 ca00 * 1,: w..'04 - 4
(4 o4 (4 40 0 I ----- 0 4 f 0 I-4 NO t1 0, 44'0404. 00 (
(4 *. 4(40.wo H - - - 0 H N Nl (4 0 C) a4C0- 4 4 4 0
0 co 0 00 -4 4H4- (4((44(44( HOmm m v: 0:0 v: V1 ( 4'4- -w v ( ntnnn(4 n
H0000 H(( HHHHI(4 4 44 44 4 .. (104,44 :
(4 *((4 -00. ('.-4.404 A-48-
0
LnH
w QI
a:0
0
0%0
0n 0
LI1 o
1.4
913
Sn +N) H
000
t% 0 c 0
H. 0 :
0 Hls 0400 P 4z TL t 0. -9 1 +
H M H ; 3 H m 4 O x Q:4m :t
Z W+ H +00a; 4 Ol 0 P 4 m 14 %
H I.4 w H::- 4 1 ' o 4HW104 - 0 +NH I-H z 9 0 4a 1
it " 9L4 0 :.-P 0LI PHL AI x~ It1 1N14H4 u Hm 64""X01 c11"%4: 9 wt
0 0 3 0.. H P Z p C.. Z g O4 Z N:ZH4 O fZ ZZ jZ Z
m o : .0 ul w+>A W PW 4H0W . 0> 0 H C4 P W WP z
O mmH 0 H LIH~ m m .4 H - 0 N.o .3 I U .+S ~ ~ 0 .LI1 :. 4 LI HO LI 14 0 N .
~0 . -40.4 H H LI 0 40S.4
o~ ~ at.. o4 N NM H .4 W IWM0 NM- of 0Wr , 0r 0 W 0 M H MMVH~r LIC.I- 4)o~a Hoa co e*.4 M MM~ Mo o4 LI o* 0 -
PIL C.0o 0 0 00 0 C 0 0 o'- 00 0 -00 00 0 04 -04 0Lo 1.LIcoo ooo o o o o cLI 0'71 cc 0S N oc oc o a Ho o o o oooo0oo
i.4.0 LI j -LI Cd ~ 4~ +A14-49
0
3,4 N
ti) - ' .1~4 H 14 i4
V 00000V C, o 0
%n HI4H 0
iN C- 4 iN
; 000
0n
*0
aV 0
.4~ 0NP~24 f4 .30)
00
* 14
v n 00 0
94w0 4 It 010a *** 4 00 0
-4~0 00 0.H H 4 .
N 1 N0
~0 in 00;1 4 4 *.4-.44 .
00 Dx 4. 000 v
V 0 00V 0
I in c4 vi 0 4 0ao 01
0 + 4 niv n-NI 00 0 A
0000 .0
0W 0 4
A~~w 0000 40ti coHH'.
.4J 0 4 - N i 00N H0 00
14 W 094~ W4 44 H
S ~1 ...- U U
0 0 0 0 0 44
Z.4 i V0 00 004
H cu0 04 000 1)0 4 0ow1
v W F4 A .4 0 n001 i-li-wH) HO N N Q 0
0 44 0 00 C . C 0 0 0c
r. 0 V c v 0 .4 0 00 00000 00104 v 00. .0 0 0 00000 00
1 ~ 00zV O 000cut>
i~i4 4. A450
U%
0
o:0
0
Sno
xux
0 w00 z
V) > -o4
C4 : 4.a4ca 4 0
C4 ~.
V.14 ( 4 0 1Z . (4 0 c
0 2:( H4 z
Wn .4 H0*i an~ 2
0 0 E, 9 0 1
m2IS( SnH
C4 a 00Z 4 I, 4(4 Zj 0 0 (.i04 0
w : cw omvvzzot2 Sn ( ;.H MZS. - 1 0E.t" V) e 61 0 0 :- a 11 1 4 H.t)1 -44
o 004 00W Z n(S.' S2,a!T: Z0 V4 -4(4:> Sn CC )U
w =w [ 0 420 * 00 0 :at0Cz z 0 U0 00(4 L~~o C6: (4 :h1. 0 -U40S 0 0
N. 0 ~ 'Unn H
1-S 0U42:02C4'~ A. 51
0:
.c
(4-4
00Ch
.4
H I
in 14.4 -4 )tHNn%
v v u t.0'4n
-V 0
Nc < ,I-
0300 * 4 a m0.44 0
33 . 0 .. .3 H3 14i0. (J.34 H 0 93 0 -. m zA. A 4 r W 1-3 In %n C
031~- N( H 9.H* .II~~~~t .4 3. .* ~ '
*10 v 0 N .- 0CO 0 HLP go> 0E4baW .
.F4. *n .3 H Q E - -413..0 cc L* '!H 03 . N -. 03 0!4 ; ; -. .C 0' 9'.1 0 1' M. Ulu, goi
b4. . a J4 2: H "-, .
H 30 r 4 0 M 94 H 14. t4-
>4~~ 4 .3 . . u4' . w7.04 A1 41 r. 1.I'- 731 a.43t . 40 f4 1f .2. 143N .:). 1.30 r*'(4-4'30 t- Z 0 0- 0..4 -4 4 P4 N
WE4 v4. , 1.3 .44 n m4. n44 -w N f -a; C4 4 C 0 H4 N V* C4 04N - w~ (7.t. 0E 1.41- Val t4 Is H 1.3 1. D.. .0..' w3
* ' ' C3.4 (a0 .4 V V 1.2 .i -r 3.3 ". - . N4 1 9 :0 .4 H I 0 C 0 V. Q 0. .'3' 0a1 HLi .4 034 C.; w 0 H ):[ .4 ' 0 .3 741 .2PV
4 4 z 01 4 * E* C- 4 *: U4 I O 4 (. HI..D.4 .0. 014 I4 zz 1 33 0 C63 z 0 *U 9L3 Z z L4 w' H0 3 *'.-4V. z '3 u' 0 C4 '3. V " '3 C4 0: u HO a'H-4- v.
'3 m 4.. 10 M 0%9 CIO3 "M wo r CIwN .4.of W-.4- *14* UN4 "m.- - ko- N .0 ." - .,DU 0 (4.4 . 0 > 4 4 C -'3I. c4 . ~ 3 *0 '340 N .H mmmmm r.. - A n43uH 40 -'H.4 00 NI3.0.... 0 . I--- *(4' HC.H.4 0 -'0 000-4 -- 00
H .' 1. I' 213 - O' . *' * 410 * ~ 1.H3' ' 4 '3.A-500212
05
0
0:
14 9
.j
CO
001
0
IPm m 14
.44V
041Q 041 I
'00.40
040 M
m.- - N00 00
C4c Z (40N0 : :9C 44 0040' It 0 0.a L x1-0
-'- .42::4 r 14 1 0 1 g4- wC H 0 'd .401-
0) E4 = 04 a M9
0' 3 EI-0 N n A9N =0 -1 4 .0.3 3=- .- g4 - 14 - f4 Ca4 S
10' CC z ' N 0 A .3 u N. 9z r1W 0 :9 z - z ' 0 f.Hc 4 .: x:02l 1.4 z In9H ( 4Z [ XE L
[4~ M 0 w40- Ca' 04f.14 '0 EH .4 1 >-9
r.4 C60 0 4' w
w La Va. Wa 0 : I 43 4t1 j9 a[(.0104a 0 0 t-a 0 0pa t 0L
+ 0 z J. - 1.4- 0 0 Z Z(z O (E 0000.1 C4% -9 N( -. to 41
.41 24 0'.n - -i W % vv 1A
MI.41 MWC ..- 1-.- U p .-4 0ct 4 0 ) I0 -0- .. 0-% . a-
lao~~ 0.1-
4~ M 044 0 C4 4 H Ha 124 0 C4 gp g 4cz 0 U. E 0 w 01'- 1-.w ( 14. (4, 0 tj4 04- Cd0 00 =C-Cxw2.4 aHar 0--I4 41 H 0-1 .4MA
C414 .4. 0 0- 0 .4 0 - 0 0 044 NN 40.0 0-. 41-4 0 -. 4 41 Ca *. 10 0
.4 10O 141' O 1'f. v -In In *0 %0l A
E. C%0-' mv k I'0 a -l m - -T. .44 w4' H1wa Il V m 0 -1 m~' vowINwm0 4Nmw r h mv0o WoW0 W wk'f .4rr ' l1" r 9w o ww00 00~ 00 0~ H.4 .2- 0 4 6 40.
o0 0 0 0 0 0 0 0 0 0 0 000 00 0.0- 0'-4 (41 01. 0. 0 b 0 0 -1 *4.4 14 "
1'a0 Ca' 000000. 4411~- 2.-.0 0 000000 00000 00000 00000
-. ~ - -- 2 0.A-53 . l- .1
0
0
Ix"
.c:>
EnoH:
co
43.
0OH . 19.- "4
.4U
EHr - ! '0
0.4 C JJC A
00) 00*
H- 0' * .0 -i *4 Ma 0 0g
0I w 0 .a0.X
0' 4H C 4Z* ga CU V f 0 CO
14 -E .4 CI*ACw 4En '-'-.04W 4 -'.0.'0
4)1 al C.4 V '0w 0. f . N .0 H0 44) ' 4- 1w
O + 0 1.4 +4 '. of ix of0 4 4G
0 wf 4 V -- -. 4 M 1441V . 1 . -Z 1 91. w 0a , E Z Z.
H% 14. 0:C
: U - :1 QW a4J 0 0 4 E4 H -F . 4I 4 4) 1 n- H C w C tt
9-C L,,4 rU P t4' 64Zp Ut .
C1 0 U0 4-E 0 0 *'H 14ZWW i4;CU 0n 0N 00 P;LICO.C 0 04 W-C *. 0 p4 0
*1*00 - U- 00aC '
H 4f 4( mmmmmMMMr- v1 v~ vN' '-E tnE CL. '.4 o.4 HnL n o w%0t oooD
41.4CC EN C C 114 14 It UI 4.4- C.4 C . O O 0
CCC ENH CnC .40 U~C - C'4C CA-541.
a
V,
4..
C
o~ku
4.4
E~E4
OHH
404
.40;
N' H 4 .
H... C,. H l HH H H HD
00H H H H
H3 " C4 H H 4 HHH-. n:HH HHH zHz z It if z x 0-1
v O.4 0 a Ht) H H H. V 4
*~~ .2) t- n
4~~ C.. w
C) C1 4 . I
C4 CA 0%
000 0 m v ow -w- , mv ,w 04Nmv0wIw00mNmv -0
mmv a,.4'mma 0 0 0 0 H 4- 4*0e0 No N2) N".N .4 . .4
cf)L.4 . 0 4 40 03
0 .- 4A-55
40 0. z u0 9. 4 (
*4 0.4U V :44 41. H H H to.
vvvvvwv~40
4t
00 '''i'0 J*tmo
('4 '4' 00 '?000
co 0000000 *4.000 0000
toO 4 000000tOn H1 A N V
4
404
A C 40 .
4000
00900
43~ ~ ~ W 1494 -3 4400.om
00000v o co cNiz ooooo 0. o *
00
0 0 Of.0.-400
.4' 00000004
IiI 4 0 0 0
A1 o'oo a tO WH N 1:
A. N4N4N040v40w4
4 Jo
W0 00 > Lf to %D k-ico
ef N 40 N 0 NI. N N
0. 2* :r a a M3 c4 :r N N N..0'- w
v) c4tW't 0 1.3 0- 4V .z auA.3 P40"Lt. -4 tO.41 ' 4 02 04 0 1 .44 40..3.3 t- 4, ?o ?.?oN
u pWHo at .4 00 0 0 4 A t t.1t~ .4 .4 w- t.4 0.E
00 000 00
u trl F. . - 0 -t N N '4 0. N.u i % mr
v o 00 o oo c o o.444 ooooooccc.-21..0.a' 00 0. e 0 0 o c oo o 44o 0 c.Z0.Z oo~~ou..2Jo-.o.-4u
04 0 C4 i 1-..4 1... 1' 4 4 ~ O 0 . 0 0
-4 0:4 404
Ha 0
0. . o 000 000 '?'?''??U0 0U ~ o',A-56?0
04 n c . f.
0 0 000.U' A 000
00 0 4 It..
0
(7, . 0 c4~oo
>- 0 000 0
0: W~ 00000 9114 *0 000000*0 00000 0
%n c400 0 0
00.-4-. - m1 V4
cc 0 0 " 91MMU 00000MC~I 0 q0 f00
ccw 0 ~00000 ~ k1
1cc V. 00000
cc~V cc000
mN 0 Vo~ 0 .
a4 1117iX14 00444 w1494 u
a4 ( - ift-
CO 0 0004
N1. 0a ta.4.4 0
r, 0 000.0. 0. 0 ,4 :0N >104 0 000~f 0 0000 z ~
'00 00000-4-4 V~t 0 0 0 0 0 4
IIt P 00 000
cc a,2 cz
A4 000 z oz t40 .44ll C4i% C4 o
w~1 1l1 4 9 1 4 4v riM M0.0 0 e I a n 01 ,4 toLi00 0 C4 z 9.0 co : 0 a N N 00 V0 0 U30.- 0 =1 C3 C4 N a4 I000000 Ek 0.. N WZ n E o~ 400000 : Lw a A N~
4~~9 O'04 0 0 0 4i4 1z0 - '0 0 0 0 cc W 4140. O H .3 1
'C 0000 0 0.a0 l 0 .4 0 W
00.-44-0.>44 0 L -Ou1 . ce0: ) E D 94P a z4- z 0Z
M'.- ;~' EI 010% N...- Ed4 Z Z U4 04 4 V
.0 00 0 too t4uA Cm0 N DZ EH O000 p .1
00 w a, 0E 0 0 *O W cc E 0w 0 W a4 9. ic-
Z C9i~4 -w .4 0 0 >I IcI.' aLi( ) %0 N'C1.-4..4 4 I-4 z00. 00.J141414 zLi 0.00 ) 0 0 0 0 W H 1.. 0 z 04U zlcc0 z H 0a 0d
04 00 0 C4 0 D ZZP H( 0 40c1 (4 14 00 !V 00000 00 00 . 0 c~c 04 wzLs 4 03 00 16 'V 0 0 1 % 44 0, OOH *-a O L 03 00 0ca 00 0 z$ 0. N ~ ow$.40Li4i N 0.0004 Li 14 Li 9. N, NzOwNNa
0r 0 01z4Caicc~~ 0 0 vL 4 L - 4 E . 4 4 4 I L
* A 0 0 0 cc 0Li Ow14 I.4A-57
0
m U
0
0 11
(h. 0 4 1. 4 C
'i' 0 4N . 0 :jL04' * 0 4 o W
0.v 'Cl I. U
0 %4--4 0 4 C
00 *U*r 0 14 0.040 C i a04 . 0 " C R
a0 a 00 00 4.0 4
0, *0'.0 4 000.04 4 UNv v = @ 0021 N0 v44 .4 0
MO 4)- 0 0000 C 4 r CM .4 0'40.U U'4 C,
t* 4..4 r.40.4'44 .
* 40 4)4.440 : 4
'.*'0 0 " 0 -'.44U
c* 00'. OU0 C II
V . a 00'A .4 0
w*.~ 'V 0 in4.0 v - V)0..00
0'I t'. 4-.44 .4V0 in 0 000l ,
U4 4.4 M c4. * 10go*4 40U04.0 4 -A40*0 A. *0 -. 4 W
ta 044 -A '444400 .04.4 .A -2 r4
404"'o U 00 ' ' 0'4 [41 0.4- V4 o u '.0 04 *O t"*4.44
V* ' 00.v 4) 0C t . 'd : 4 0
4 4 tp 4 0 . C 0 0 0...4 Aj00 m > 0.0p44'''k
-. 00 44. 0.J 0 ' 0) C a.* 0 0 . 4A 0. 0 V'0
0 004*4r 440- 0 0' U U0
04~0 '-A4 w04'l 0 0 4U*0'X40 0 0 '.4 0 4 U. .0
N4 V) t ' at.0. . 0 0.4Nm4L ' w( 0 t4 (I D.4J04(h4* 0 0 0 0 00 J.04JU *. -4 '.4 N N ('. 4 m "m4 N
00000JO 0 00 00.' 00 I '4'.4 '4 0
000 00 0 00000000 -44 0 040000 0
0' '. 04 0' U.4 U.4.-'.4 A-584
0
* 0:
0o U
0 014 4
14 z
11 0C4 90 z -0
0
0 6
An 0
C, l 0.
C44C
Z1144.4
00
OHH
A4 00V N
0 0 0 1
n 0 .4
*'~~C4 x 0
0A to0H00
0.:1 4 '
o4 w.4 0 4141 V 4 0 400 0 0 0)4 6
040 0a '0 0400
-1 Z '0 0WWEaEca * 0 C
00 . 140 . 00,: 149 .4 1 014
04 0.0V Z0 % Cz4. '! .
0.4.4 , 0 U.
0 H0 w 0 x 0 014. 14 g4 H4zzJf
0 z00 0 W 0 0% H '. m 0z tZ4 c4 *. 4 . 0 H .
. 00 0) a 14 . 4 "1
4~~ .30 0 0 a 0 p10 14
a, 04 03 t4 u 0 w ,1L. 0 14--0.40(-
m4 ov4 64 04 14X0 IN Ham
4.' -. 0 00 40 ZO 00 40 00w" 0 W0 0 0 I" OZHoz14 1
0 L04 1 ) " 1-0 0 0141 0v0 ( 10 0 W 0 -: 00.H4 0 4H
Z ! 0 0 . 0 00 14 o V.~1 o C40~zp Z E4 0 V. o o 0 V 0 0(0- w w wz 400
V C4 0 v .4 0 0 0 14uf zz.'A 04 C4. M co 0wo m4 0 4.0 H>. 1 4(.10~
0A-591.. 1
C
0:
t- 0
00
u N
00
a oC4 c 0,p9Lq4-I 41.1I
1-0
~A-60
tnH
a: 00
tn n
.. f , 0I
94"t r-i
In coIDv ch 100 H Ch -I
.44 (.I .4 I. -1 NII0 .4 vi .
00.i N w w4..0
.4 ;; .;4 %o1 .
lo I iI. N10 v m
0; 0 co-.f 0 00 0 .74
lo w. N N4 1* ; . . .
0 ~ 0 a 0.1- v). .4 Io~~c o 1o 7N
H! el. c4- v4 q cn c. N03 m3 r4,o4 O0. N7. qq 01.7oo N4
0. u NN 1- x44 C6 U.7A ; coco4N.7In r.0 ~ c w *-O3 .. *
. 0 UU N0 1.7..
Z; H . w. H . . ..- .0 ... ....000 0 II
H ; P t H 0 N m~n0 0 .4 4 .7 : *M
* ~ ~ ~ . v z4 00 ~ 0 0 0 4 1.* I M33 .N7-0
H. Mm. (d 940 H4 -4 r.. o~-4 a0 0 00 3 o.- o0 n01.-c 0 ~ r Z 0 LiZ %,o0 m,-N 00 0 0
H~~~ N.~( N 44 H t4.. e! w!H v m o mO O.O.M1. .4*... .~ ~ H '.Io o40 oo 0 . 14-i I 0
u w1- w 0 ZHO -4 , 14 0 .4 * r. N0N000V0N 11 loM. . . . . . . . .
Z~ ~ 00 . 4 '1H 4 F-' H0 , F. . 4 1.-N 47It 3 m003 . . . .
z a 1 w ta v* v.E3. :i m .a,* 0 * 3.. NNOO O . w H07n t It
C40 - 10 %0 [g 51.. 00 P,. X...4-. 4oIN03Mq.7.en.o.MXZ .770 Z0 Z.... ~bin03 0 4 H O.....N.44*III 0
o 00 0 O 000 00..0 wH 3 0 o 0t
Lb 0cHI5 w 0 I w~ t-- 0 0 00 00 00 0 00m*(14 1 w [ 4 Z. V H M. 4 H 4 NM7I03 4 .4 LA NM(40 0 04=1 'aZ0 3U aW 0=H1 -1 4 44 (A0c
000 00 H0) f 41 0 0al a
A-61
0:
00
0%
x 0
A
v I
244+
E. .
0 4-CW
14f4-
'Cma
00 444 'vw = ;H
110 0 41 + 3c[4 1 4: - .4[4 =9 C7 42 C4 4 t
ca *eC "4 0 wHH ' 00 00 !t7 C U E4 N.[ C 3Z a, m N au aua'
wN zI. HHM4 +04 0 C4 Z e . .4 0II *u gm..u .42 £4 V.
no. * e 4 . -CU o
H ;C Z*v 4u C 40H.4
* -CC 4 '-C*ae. 4U4 ~ -* m a, HE4ooc ooo
* 4)4 404 4 AH62
mU
:>
940
0j
coU
LA H
HaH
a o
u HH
401
00 'A;r
N Nt ,
04 C4p (
0.4444004
CO 00 OH1 o o o04
t0 n t a a aHOU d: c* L i
0 0 2U
H H H
Q 1 E 4 :r f4'74ZA(4E XXXZ0:
0W 0 . H -C4 0 94
4 4 C40 W
0 * ')UI
'.I 0 4 04 * H4 '
040 N H H-*3
~0X 4 . >.lUx
a: H.4 v w go
co0
c o 0
ul 14N C4 in 1
o:00 m M O
D:)Z X 0000
a4. r, o H . X
a: ~ ~ ~ ~ * XOc %z o.H o%D4 4 0000-w4 Nz N NM N
QQ0O
co 000000 00MCI 000 00 0 ccJ
4.2 00 0 000
1- 004
.4.3444 :.1 444
4044 04
m. a a N Wz:00:0 F14 N1
WWW1000> OMHOM
00. a.NmNN.
0400 0 to00000
04 94 n NN
.0 000 = .4OHOO. 00 m.0 UUU to .3014 .3 p 0 F-z I- 0 2w A p "39
a4 UUO 0 -W 4 9 4 U 0 o H o o~
do4 0..00 0 ~ 0 u . 03 30 0, 42430 C 4 A v =0 z . 4"H " 4X z NN m u v~u a HN HHW
0 F4 0
*) C, 041 40 00 .440 00V0 U 0u 0 In- w w 4010 mf.4 0 NNN(NN4w0 40 M040 %4040N00000I m 0.20 . o- 0-1 V4 a4 1o MM MM -- F00 MMNsI.4000. 04l
H 0 V 000000 a 4414414 N .-4 .-4 MN04un 0 L, 0 U 000000 .0 0 0 0 0 0 0 0 00 0
Z r. 00f. F4 -0 0 a 000000 v0 0 0 0 0 00 0 0 0 0 0Ic~~ ~ 000000n V 0
z4 4 0 0 0
z 0 p 0 W In 4f N N m
z -'04 .
'A-6
.0 0
0
!0 000
0:0
.4 0
0 m
00
.0 0,
4C
0 N )4.(4 [a0
ca~' 03 Z0H 9 00
N~. 0 (
00 0 0a Z 00
A-6
0
0
tnH
9400 1
94.r
0~.
OH
C4 .4039
0 0V) 9.
NI-
o Do,
000
.4W-.
0 : # 00I 0 .4
[4t4( 0
a: .4 .
1 4 z .Dw u)z 4 00. H0ov
Iz 941 'n 49 0 3 -NH=
* 14t131- 0ca* 0 z X.0 000 0: 403 HZu
*~ U 0 H0
J4 0V6 003H034 0*4 C6 H~ 0 1
to 00.4 0
* 2 0 0-04g0Z 4 zH* 0 .4 03 0,0 U) (
U 3 "I . U).01'O04.3H*H p 1 Do10.0.40103 H
'a* 0 0 0 1,11H U)
* 0l 0 1 3011 10z- 3 0. DU 0*4 0 0 11 M-0U 44 01 00 0*4 m3 w U.o t z 4 ( 0 0 14, ft
g =W'0..O H HXU )) H0 t.r
z o) .440.H0300H o .4V~d*0 Hi N.00 H % .
00 0 . 0'~)4000 v 0~13 3~
* 03 I~034.J0A-N66
U- 0
C41%
0U
ICCLtn 4,
.41
0 k 414 4 t e * 4444 444 444 44
'C 4 4) a4)0c0
414' .4 V Q. 01 Ai 4.
*4 '. U 1 ty 4j 4o~ c C 01.. A 4
-4- 0*4 .041 4. 4y a 43
01 c 6 0 0 41 4.4I0 4-4 41 r-4444 4 0 0 t.-4
00 00 Q 4 010 t, u4 0 -0
'1. 00 4 .. 04 0L1 WeL 4d -.4 C: U mCtI A. 4j 444104 00 - .4
00 'LI 44~~~10 0 00 4 444
:3 4.4 0 N 04404 N1 N Hi LI 41 4
0' 0 0 c 41' C *.4AA..40 v4E - 0 0 * ) 41 VI 0 0 4 4 0' I01 L 4 4h r. r. 0 00 rt 001
*~~~~~ ui w 0-4 0IL 0I u 1W
>4 0 > 444.- .- 0 4 .4'$ > 0-A4 d4 " "I 10'4 44 4 0 4 4 '44 (A C4 41 404 0444A I
c 41 * 0v W00 10 1 4 4 0(14 M M 0 A 444 C: in 00 00 0 W4 4j 41 41 ' C -A44
4- i . 4 4. 4 44 . 4 .40 410 In4 0
944 fC * 4 0 LLLLL 0 41 4 44 *0 .4
a041 -4 4L-44n 14 * 4 0 . 4141 44 0' Q.4 444 44 4
p44 'a 4i ,.0 0 4 0I0 '. 4 0 0 0 44 ' 41.
a 0I *044 4 40
L* u4 LI 0I 4 41400N '' U 4 W.0 44 I 0*Ap a 'LI 0 U.4..4 M ... 4 cI14 4 0 44
0 0 t'4 A4~ W. 4, U10 t, 0- .4 '
C' w V).0 ' 0 .3I
40.0 *o001'2 * 4 .0 4444 4444 0.U uv
'444 VIO'C14 4 04 44 ~ 4 A-6074
0'0
dl.3
61 0 N.~
H0 co
.0 co
0 coo
0
0; 0 00 6 0
0 P: 4 0004
4000z zl
944 H4 Wco
4 00
400
N V. 0.0
.- 4 V44 V4 0
NMz 000 0. 0 P
A14 V. 01 00u
A44 0 00
0 00 W) 000VI A V 000t C
-f 'omfnt 000 I*- - 000 N
01 0 41 co
II 0 M0 0o
:> N 4. N 4
14j .0 0I~ 0
* N 0 JJ A....
-4 M * 4 X
411440 0
N0 r V H
+ 04 a444 0400.
04 0 (A
44 4 0 " 0.I".00 *4 . *Z 4140 %1 C00
u4 44 0 M4 4H 0 000 co1-4 M4 04 0 a 000
a 00 44 C6 uU. 000
C6V .4 04 'ac 000VsN 0.1 t* 0. 4 4 0o0
w 0 a 4 o- 4i 4 %o1. tIcIo tIm- n w( o 41Ln..4 o %oooo~w , *.- .-4 .4r r .4.-4 N N No CO1
N 144+M~0 W0 >.. 4
4444.4 'A-68
0'0
4.
0
1.3a
0 0
01.2
x wco
H4tA .,a:
094
p41E4
0~ 04 19-S .0 0
4141 4.2 .-
t4j 42 ,-*
00 90 00SO 5' 04 m
10 '0 04 0V0 0
0
gI .: 0
0: 4 z 1w. wg 342
U30 % 4 40u 0
4, IL2
H..0 0 0 j H0 WLi tz2 03 Zz (Z H3
.4 3U -2 -H41, .41- 0 >4 N-9 aX
n2 0 0 H 0 % 14 2 H
0' 0a Li . H H
n0 0 N 0 X1.3O0. 12ta: 0. 0. * 0 .ZOO -
w ow Liuz of IHO 1.2.
VI ~ ~ ~ X ZN .2:040 o f z [uIL7)-0 z~ 0 :
0..4 z zN.4 0 u
0 -' 0 D t:0. zfqU4 w 1W 4V 0 f 0 Z E 0 0-4 zr001 .40:
0 1.. H 1. = Lw= - 0t-0 H0 H9z. U) v z : o .Ua w. :..2:WU 9) 4242
m2 0 0:Z 0 0 0 0.2 Z: 0
0: o 1.3 v U 4 L 22 4
'2: 0 0: 0.N. '2U~i ~ A-H69
00
z
00
41
m 4001
HH
001OSU a;C
CIO
0100
4. 9
1400
404.
usl~ .01 0
X4- 9 41%414*0 .0 *4
r n.* 14 X1- 1410 0
050le oi~ 0 41
050 004- 04 :1 00j C
m-. .0. 04 -4 40 0 n %
00 0 05 545 5 11
09, * .0 . . . 0 . .1.1....
3. 35 '.. 2t D, Dcs '0 .Di:r3 1o 0 '. 0 .oo0 0 0-k
r- r 045V004560 W6
a VA45 0 0 W,...4 w5L '4 .00 111
00 * 11 '0 1 100 145.4 1'a 'Il 1 '04 . 4.)~..0J *1,.01
* 0 00 to .4 -4 4 V) 00
*0_5 C4 '100 I.000I
41.4 0 p,044i1.46S
04 0 , : c C IV
*00 0 0000000,000 a.OO0 05 0 0 0 0 001 C4
0 0
H 10.450. * 000 0 0 0
*4 00 0
'0~1 0.00000000 0g. 0. 4 F 4
* 00000000000000000 ix
00000040 00000000000014
0545A.70
t~0
-0.
z
94
0.0
143
4 Lt.01
%00
4.
T 1W010
010
0.4
x£44
z 0
o~ '0-4-4
000
.4.
-. 041"
0 14 -4
94 N .0.4Z
04 m£ 4 z9 1 z
z .'a - 01E4 0H wot4 g 4 H £
1% 0 0 u X£ 90 4.O -Z-
94z If* 4C C4 g0 0 01CW =01H£ H .4g
ca 0 v z W0 00 z 4I 0a20 01 yH -; V o~ 010.w U
4m 0 H.0 01E4 .4 4Z ZA - z (h 0£4 a k. 0 l u .3r 4HC
0 2 w 00 H 0%, w 000 j 0 ~0 14 IZ 40 V Z 4>( N 0 1 C4 0
Z4 W Z-Z. -tI 0
(4 O440.HZO O .4N04H1 94 ££0 HON £4 * *4fl0
.4 01£4ZZUI-.Z In 01.3 0 0:- " 0 0
z Z44.0 O 0.H N. E.
z0 . 1 0 :0: 2£4Wh4 l 0
'A I- W00E041zC.400 0 .3 Me1. 0. O= 0UZZ0(.. HO w EA aCH 0 0O~wOH.--H 4 C6'( LI -I-z W 0 C401 Ul0HI0 .a0l N 0 0
00 v .0 u0-.. 4I'
H 0 0 ~ ~ H 0 4 A-71
40
00 0
U 00
N 00
*H N041 N
4 t
0:0
F00a: 000
a, a, 00(4w oChQ:i 0 00i o
1 no -.400
4 0 ., + L%4. A d j9
fn 0400 v4V41 'A44L - 4E
4.3 41 1fl."04~ 040:+0: H
141 , r4wW
ON m 0 w
rnj. .0 V :0 ul
.4. A 41 WI + M 00*4404 0 "1 to 1 v 000
00 *_ a3 03 a .40 000130 S* . m0 u4 L, W 000
040 V0 0 CIl 0 '0c 000.44 OU 0 C 4.. 0 00
00~ ~~ C >. ' 10MOh ~~1 (11- 0 114
I * 0 4 Is0-0 C U g4 m . .. 3
H 004* .4) o- 1n A4 .444% I 00
.0* co V * 4 co m 0..W
IC 140 .4 .* 0 12
0.0.1 0 4 10 00
IC fl 1 ~ 41 0 qL .
IC~ 0 NU. -V 4U 4j 00 A m 410S
*144C4 GOC 0 '1 9 0a0
:* 0 4 4C 4 0. 0 4*NC013 UU N a,*
NC ~ ~ ~ ~ ~ E 410 C 3 -l.00
41*~~~ 000001 .0-* UVU 10) Ma 41 V 0 ONoo
94. 00 " cw V 0 A 4 000.- 4 41444 0 .4 0 14 000
* .2: .40 "* A A 000r* 4 U 3 MU UA Uo '0 000cmvtnor #r -
0 0 041* 0 41 41 0 0 4 000NNN 4N'0000 .00 0041 00 400 41000 a
00000. 0 4000 0 0 0 00 0 0 0 a, > .
41* U00 04* ~ I.IN004 A0.72
00
00
o'U
94 0
N v:3
000
640[a zC
0 a 1.3z
w- 0< 03Ca ga
C4 4 H
4 94 0 .3 v.3 0~ .2 D
o 0 *0 -.t 9
z~~C H 44 tc4 m M 0 0Z 0 .o '
(4~C~a- 0 0 0 :0I
Z ~ u) 0.3C 6a: ~ ~ 1 \0~ :; U pto1; 1
0 H 4 . z u .. m com cH wa co 0 0N
Ca 0 0 00 C >I ~ Ca UU0. 0o Z Ca3 a C4 00 N
N Z~' U a a H 0.ca Ca .. 0 H0 H .4 P wCa Hj .000HO'.. Ca
.3 0 >IV.a 11 11 0 #4 ) 4 .-Z. Z WU- H , -0Cn0C4 a F 4
It 4 0 : Z> 0 cU
Ca~r Ca Cd uV. 0 4 0 t0C0 HtIH 4 .4 H A ..
4 0Ca wN : ~ H :r= zC. z a (4 tZa 0 (4 W0 0Z-40.O00U .3 91 .4ll
X 0 .w. 0 r. 0 0 C. 0 HICO (4 0 10za 0 0IOO1.4 .-'-'H ~ N~E4i Ca doff
z zA w4 04 00'0C C34H~0 '
0 4~ 0
0 t3 0'~ A-73
0. 0
P: 00.0
I.Of
00
4W
0~u1*:4
t, A
0 0.4
5,14
tp0
MN *. 4111
4,.- 4 0.0 N ,I ,-
000 a
0501L 40 4 4
OCa 00 0 :m~ 01 0 *i V. C
O'*, c . 0 *0 0.
II v vu *v 0.C'4
M*7 VOW4 W00C" *rO t CM )O l6 1 M nDlC %
"4 C C 4 A-74
00
r .C4
> f
4 4:0 44
0 4 00a
x .3>(0o01.3
.4000
Ow (7 . wr.900 c
00
4fl00 .3(0(
940 H
AO 000 a 00 w a 9v u v ~ ~ 0H 0 co 4 c
HI .3 00.4 no
II 0
0j 1.124 m 0 Zn* o o~ 4.
vi N I0 o c4 o o
N 0 0.I..1
o W.3E 4. m1
4~~~ 0..0 0 '12. oC 4 0
c4 H (4 E. .4 14
00
0 0 0. E. NNA5 0 0 44
a a 4 .4 '0 00 0 W 0 I oL0 : r 0 0 3 ' 000 0 .0 o 05 00 .-4 V 94. oo U
III .4Z> o 3 4 o4 P 0z~~1 C)c 4L
'0 0 o .'0 0 41.3 > 0 0
0 A-75
64
V-
z
w,
0H
in N4
14 r
14
.4 0WI4W410
Ifl Nz 0
or-o*.4 9.000
a-4-04
0-4 0
14Z0,N 4 H
0 ; 14S 14 0
Z :1 H si14 J H4
m ' )414a~. 4 00 0 = 0 : In0 9. U
4 u n o0 -H 40.OZ4 w f.
H 1a]00 m0 W
> : N 0 40 Z E C4z 4 0
m 0 00U1 .3CV) .4 v - I 14H a t Z. 14 w 1 4 P x
H H :300 .0.41 nu H L94 a, Imu !!0.4 14 FH 0 CO 3 0 00 H oWV0 0 a01 H ~ F4 Vu *0 V Z W 0 1 t) 0 ?U ? V. .N, *'U.40-14 - o 0P.N CZ4U4140 4v H 10.14 xo " M 44" HH U*.000 -oN 14(. 4 *0
.4 9 = w U14Z 0 t-0 H0 IM VN o U H4 -no~v ~ o 14 401
0 * .eH N, 11 4"f0 0 1~114l114 N v~104444 1
H 4-. OO140~.4CUA-76
04
0 .0I ..
V)
p0:
010
0 A0
01*
>.$
04 14
('4 43 t4. >11
1.44 4. 4 y
14 4t >t
w0 4>0 a 4U
0 4 4
00 0-61 N01.0 >,4~ 0 0 : -
.010 tow .0 44 . H,'" 0 4 .. 4j N
g L. 04 a
4,1010 u4.0 4 0 4 4:1 .4 0 1 44 0 0
3t' l1 'd -.4 > 44 0 0.('4I V0 U. 1.4 4 1. + - A
N 1140 0 .4 ' 0 41 4 w444 c') 0, 4 0 41 143 10 d 0
U.'0. v 41 1. 00 4w44. .0 . . 4j >
-0141 1 0 5)L4
* .4 V) 0 01 4 * 4a0 4wC:1i' 44 '0 a,0 1 i
w44 00 01.44 14 T4 t414) Q).4 0 > 0 4 !j 4)H
040.44 '144 0 0
Id44 '0.40 V 'd 4;. u V.4 4 w. 004.. '4 '. ) C4 >
4*D'ti 0 4 0' 4 0..
41 V >' 00 a'4 4 1 4, 0 %4u w 01.. 004 .4 a -
).404). L04 . 44 .0 ('
44"404 4114 -A 0 4 04 010, 1.0440I
14 4 1 000.. 0 40) '4 .0O4*0' 0'D' w 4 4 V I4 C L'04 W 'd0' u - -4 44 4 L,*0 0 s M 0.4 a a4. - *40 0 F!43
w44 4 ' 00 0 . 0.* 04.4 01. 0
414.00 '4 4., Ij I '4 . 44 ) * 4
~1 .4 q4 *4 4 N4. 44 4 4 N4 N 4 m m. m -' m4q14mm 4 4. v v40 vWmLnomi
144.1141 '0141. .. 4 4 40 4 4 )1 4 40 4A-774.
C4,.
'. 0
00
A 00.
4 0 V
H 4
6 00
u12/ z 040ofH ONa 0
0~00
.49 00 14 4
cc1. 444 0 co040 . 0120
pi.to 4 A Ioo
00
004. 000.40 .3. 0.
. P: 0: 2 o 00
00 0002 a 000v40 00 V . 0004 0
ZI H 0004 LnQ
IcI
VI
000 04
coo N 0
~2 0000 0 0 0 U0
2/0 HH a
a 0 W.4o 04~ 94 z
V 00 W4 0
24 2
~0 z U 0 2
31 0
0 24 I-E0 Is 0W 0 0.1.2 044 0 C
V.4 n 02 0 c C
0 42 24 4 042 w 0 w Hi a 0 .0 0 .3 H 0 0 o 040 .4 01u0 2 0 0v .0
a: V0 . 1- 4...00 .
0 0/ 0 0 . *4 0. go 0 0 z> C, 44 2';
H 0 0..>
2.. 0 2/2 0 0 A-784
00
c0
U
P040
v 9
0~ 4Oi C4.410
=00o[aoz
.4 0
z 04
w- 0 .3z
0 ZN " 0Zl -4 a
N Z zO1li 1 -.4 N 14
> 9.L1 1.2 40
O a Z 4 9D P 0.
Z N o 0c0 0 0 4al .
U .3 0 0 0
w 4 V to 0 414 0v .1 0 0
wn - O.4 U 6".., E o !" w to
a'4 0 0C.U 4pG1 .n f-Z ozZZ I
0 0.1 1 .11 0
D OUUzwQz0 Zw 0 00a m 11 ~ IZ 1. z UU. :
:4 .4 L 0 0 4!C"1N 24II0 N Z l Z Ec
4 .4 c412ZZU4Zlo~ H c -6 ty112
= 0 uQ4I m~vwz4A 04 :1U -0ta z 1~4Z X W0 WN .
> 0 0IA0.I211 .1. N . IO4AA 1
I' 14 200100?.A-79
4'
.4
'4: 0
49
0.0
ina
14 P
a:1.33
3, G 00
3(00
ON9 4 9
:3 z. 0 wAn' 00 0 00 E
4). 4 '3' *M 03 H 143.
04 0 . 4 x.31.304 .414 4
00CII >40
oo *.' >0:30 u0 4 ..0"D 4 0 E4 X4 01.3
w''. 0 34 1.4a 1* 4 * 00 4 - "a' . N co H) 0. 1.
41.34 4 C4 1.1. w~ I
4, -3 0 0 U U 1.E4 .3 .4 03 0010 00 U4 '4 C4 '
'.4~~~ 004 :0 N*, '4* 0' w 14 H .4w 4 Z
4.0O a r3 0 14 .4 H 0'
*1 *z 0 C; U4"03 X*13'4 4 0' ifw M 0b 1.3 Z IN4j 13 00 0 a P: w 04 0 W
40~w 04 03 21 ' X040~~~ A3 00 :N
V'4 4 1 '0 0 0 '4 u04 04340 0 1'N 0 j H 0.~. W
o. w3- o 4 04 0 0 0-0. 9r4 0* 30 w '4 0 0 .4 to~034 0. 0~ 0 C. Z. 14 4 .3HU3. 1 :t) a 0..' 4j 4j V) w r z .~r 4 014 0 4 v. m. w 1V0 f % 00Z ww40V40 * c .4 0 03 n c W 6 %X ..4 'Az3 N' ev g ..31. 1.3 1H
03 t, p-'-4 u c0u. 0w I0
'4 '4 4 0.4zOa 0' C4
*4034140).Q333H 4) 0 '4 0 c . 11 1z -4~ z 030 O 4,04. JJJ w3 '3 4 3 '3 .4 (4 wou~'am" 4
'd~.1 0 VI 30*0 0 m 1. f w. to 03411 Z 0 ixV.t . 34) 0 0 0 14 1.. o4 c3 41. 44a. N 4 =WW'332~ ~ 14( 0 '3 a a3 V 0 .3 3I4 e
000 0 00 0 0 13 '4 > . . O30Z~1. 1
* 1....4.H 4,H 0 33OH .- HA1480
*-0
4
e0
H
nn
Z n
00
92 °
OH
c C:
0 0 0
0 0
oco
I I
00 wnt-4
o 0U
I-.
t. a t0 €"
I,2 U 0
I 0 s-I 0-04
430 0slAO U
A-81
11.0
40
23
01
,1
0
0f H
99%4m 00
MW 00WWNW4 W W
0flt 4 4 00
'n-I9 44 W 0
* 4 0 411
0'O~~~. 4I0 3.3.31.. 4.-4 C u 101
U0 1440 .10 t;. 4 X, 4.4 . . > 4 - > - X 141WW) 4 40 4 1 - 0 " 0 L
,W 00 J. 0 .3 4 .. 0'1 1 0.4 '31 - . "I o* u0 U
m~. * l 4 XU, .44 . > UUU z44 > o 4v-4. 0
U*l 414 le 0'4300'43v0'4
* 0 o4 1 O 1 0 0 4 14 41
40* 4 l.0 1 140. 43
0 4 '0 '.100 30411il 0 4 4 40 .4 -. 404I 40 .0 000 %V 0 -4- - 4 -1 04A644 - '4, O t
4.,0 4 W04IdW 4 444 43WH
.*0 a
* 0. 0 04 4 .. 4. 0 001 0
:0 :01 :0 *444444***** 0. 0.>0 3 0
4 cId
. 4 0 00 00 0,, 4 00 00 00 0
C 0.* 4 Mq1 4
0' 00 . . . . . . .. . . . . . 0
4~0 0.0' 4 4 4 4 4 4 4 0
0400 u 44 00 14 z04 '40 4 0 V.1
~4.I0a-- - - 0 0 '0 0 0
li0. v 00 00 00 0 430 .0 0 04
Z" 0 E.& C
to 0 w4 'C. 10 w3 0 ut)UU t) 1 141.0 43 I 0 0 14 1 30
C6 . t 4 ) Id 0 1
A-820 0 3 . 0
0'.0
(4C.0
' 4
~1
:1
*44 C
U20.4
z 0o
004
0
ii ~ .4
0 z0 04
ul >: H
=: .N H C
C4 -, H: in
0 : 0 .14 0.O 0 42: 0:
02H21 H4w4k.4L
n z . C4 u&I WM D 0 z 0 0 0
0 0 =I c0.4 w
04 0- D V-01 C42: .4 -- 202:2:2:2 014 : x. 04 V t V)I-. OU0.0.4O 2: H n
O 1 14 - .. :1 u- . n .a
0 0: J% Z 4 : 1w 2: zz .4 P4.1.2 1 V
0 m 0 ZO w0 w 144E.
0 ca z4~04 0 ~ wC0: - 0. 0 0 54 Z a
z 0 - a 01 0 H
It M Z -H Z- H' .3 0 l 'a m
H 1.00 -HO 1.2~z-4 0 OlIt 4 =1.2:2:U42: 0o
o X 4:.. 0 U~O3V z~w ):Z4 U: N0 H ,a 3a
A. 0 0 120444- 0:0 4 N :4:101011 -
H 14 0014O 2:.4 UH 4A-83.
04 04
0H
0 4
0
0.4 "
'-41
0*0to 4
0 0
*1" -4X2 1 9X 9 -
040 '7 4
-4. L, -4 p4. 0 : $ y ~A0 V- 14 4) 4114u
Ai4 ;. V~ NNn11 L0I 94 .. 1 4 .444
0) 4 j. 4 .6 'A41 A4
1doU - - - ,j 'A .4 :1 w0
U 43 09 :w owl4~C
40 .' .4 .. *. 0 a-Ia '0.. *. Is I. 43d rL :90
*.. a, 00 0 0 044 0~ cro 1.4 'lo 0 0 434 0. .4
-~ ~ 00itO 4 1
*n w. 0 '0- n0 wc 0 4 r4e 4 -w nt -wo0v Of 0 0 0 'n3ry , . - 031 3 41 0
0 ~ i U 40 * ~4 A-84 ~
0
4~ -4
00
0 Z P41141
U 0
4 00
F4 t4 00
00 0
00294 4
0.2 00. 0
z0 z 0 co
I I01414V,14 4 141
1 1 00 0:0. *z D .4
.3.4.4 00 9H4I 4 .006 00.I 1
04 W 0.-wNv N0. 0: 94 4 000 X00 z z 0 00 0
A0 000 V) 000 0 0 00to0 0 4z
00NN 14 .00 -C z 0
mi 0 .3 z
41 I .4i 40 1 02I000 V4 00 4 z
.0 00 P.140 144 04.2 . 14 14"M P.0
00 pCCI 02 2 :214S0 00.5 L 0 0 23 00 V W
.0 0 0 0.4 U A -Z 0 z M0.4~ 90 CE4 0 : ; w - H
U~~9 9 ; 0 9 =44 ~ 1
41 I00 U3 0i Z 2 .4 4 44x 01, .444 0 0 wa 0 .1w,214 14
0C f...0~ *'' ' 0 IC 241.10. P
E4** . 0 V . 0 44N ZOU U39.3 HC-I 9. m0 0 j 40 '1
14 -4 14 .4 9 9 -0.I .4ma r.l a' 0 N IV- C I. 4
14 . z1 40 4V 4. 00 ICH4 0z 0 000 0 :F4a .4I N .4
0 000 .4 j4 1224 0 44; 000 W 0 9 Z XOU.4 E.
'0 00 .0 a4 14 z0.wul f xC0 00 4 0 n 10 0-0HO004 IC
o~~I f4 . 0. U4 4 OM U Z 9 02Oa'N 0 0 m z, 0 " ;IC1".I 1
a,000 0 2 4)0 0> .4 aC FCO 4 H
o4m N4 4 1.2 4 e 4 4 Z .- .21.21.2 ww~~ 0C AC z m 4 0 = 01.-2UO .DIC0 o
U) 01 1.2 0 0 00 C I0 1.2 C fi .; Z 10Q0 0 Q. 0 w ,.~02 0 .0z 0 HH 4 1
1- a v 9 . 0 V 0 to a C a 0 0 .4 0 z Z -- 0.4-02 00g ~ ~ 1 z. f. 4 0 4 4 0 0 I4C 1 0. 9 =WWCU0200 4
0 I14II I U . 14 z 0 0.40Z C 1..u z0 14 C4 0 Z2 E4 = 2 w 4C.~ U4 .w z 0C124 .O 0 H ' .4> C L 4 IC.uU4
4 A ~ A 0 4 44 L2A-85C
0
0I
0
U
H
U
oU
UH
P 4
olr
OH
.-4
NWD
00
0 W
04 0 V
U U
0 0
04 co
011II
00
A -86
U * U
.4,4.. 4
I4 -. 4 e' 0-.
.4HO U
H-8
C4 f
ix00 H0
Ch0%
00
.4
14
41 w
D,414
'III
HWH
000
VI N j4
0> 0 10 .1 1 4 M4)H 0O .0 . U. - -U "
MN 14 1 u 10-
H V di 0 - 0 + V; 14 4V) -) -N IV WU !! 1.!
u u u -to 0N 00Dc -4 .* '' 2! HI ANv14 l~'., 4 An 0. A OV a' 0. 0- 0 .
I Ia: 0 -.- O 4 0 v 00 uSI" "0 W 4.jVV t W00 L' 4 o( u 0 - - W
'n 0,4 .0 1 M M uI S
u 14 a 0 SII 4 N 04
4J > M al :3*-:14 4 v >- ) H000Nmv 1 -W .*' Iv-" m Sm
0 ... 14 u I0 1Id d 6 AA AAAN I I I
V 0 . . . M' .wV 4 0.. N4N W 4 OU 1 W0 0fO M )N 0Sa H 0CSI NN 14 . WL 0 D U vu 0 t 0 U I I ) 0 4 1V ..4
SIIS 0 000 . -00 HH 004 HHH 14. NNNNNNmmmm( c nmc wvvvv ,L ni
if 9114 I SI0 ON OM ' M 'A-MM7
0.
0 N .4 IV I
> 14144 4z 0 M444044 14
H 14
4900
H 0 0000000
140000 0muc ~ c ocoo
Co cc 000043. 0000000
44
64 4
401.'.
44 E4 4 0 3c
P;i 14444
00 000x1
NO 4 0 00 0
NN 00 4 000444
0 0 44 14 f IN 4WIc 4c
.44. 0. -c ~U
.3 .3 w 00 0 0
0c 141 sc 00 I00 z'.000
A00 P P4 f4N NN0 0 4.4 4u 4 0 I 4 z -%
00 1c 4w4. 0.
4. 444 I0 14.3 14 HS44 0 011041 . 4 1 .
.0~~~ 00 x 44444 0 41Coc 9 000H 4 H Z (a 2000090 00 4, a4 H0
4'0 0g HH ix .3,41400 0 0.0 0 al 4 343
0. ccoo. 0 44 U - W a
N14 4r~ ,N .l g-.44 .II 04 X 4, 4, 044 1% .x
S~ 44 114 000 4 04 c
to. 4w zx 000 0 44 410. 0 0 0 0 =04W1 0z H00 00000 0 40." g> 0 0 0 0 z -U.03 W 0
0 000 00 10 0.0014g w40III0I U .4N -4444N Z H .4 ew
4 4:,,z 4 V4 400.4 Hoo o w
: m up P X1, 4. u m-- 1.14 4 n.4 16 0>04 * 00 4 w U
0 0vu p C4041.34 Z 4 %0 H0(a t4000 4 460 '
4, 14 N. 4440 0 4 H 00 H4 14 00.44 44w 1. 4,
0 u P W.N. M4 4 O UZ O .44. 0 P4 4 0
IV (V44 0 H 0 . 0 Z m--HON.~ 'A a o019F4 , QO U ~ v 0 0 >I z101144 .4w4C44-0144 4V) PIt0Jfm v .4 44 Z Z444441 0 144441 C 0 0 jP Z [0> W H 4 V) 400004 Of z 04 444 A D H 4 u
4 A. w w 4 0 04a1.N 414 g 4 " 1P4 q0 ..0000 00 ZZP~. 1.3W N. Z.4j0 0 4 14V0 4 U 10 U 0 14 14 1. 0. 0 00 40 . 140 1 *
H 4 4, H 0 c V0 H.v .Hw r 1404 0 64 m a10S4Z1 4,P lI 40 < 00 C 0 0U0~ CC 0 r 0 w =p01 fl140H 0 40 m . o0 0 4 14 4 0 4% 4O4. 3 4 94 440 40 0cc1 4 a *1. aN .w 2 z 370 3 0 00 0 0 N -. U444 0 0
H 0 00 > 0 u.0 13 ' 0 0 0 G00 .4 4 l 4 .41..
0 4 U 0 ' 0 ' 0 .2 '0 00 00 0 4 0 H4 41 441A-8148
C-
le~
u.x.
4o
01'
lA
a 10
.4c +4
00u
C0 U ! Ij N e
N Nt4 41 wA H A1
I :a - V 40 .
W 33.0 " Af 0 0-e q41 uI U v-
Is A333 41 I -U In-r4) u0 U a-.C
!3 w ~ -4 N + - c33.0 Vf -J if VM -'-
3333 4,4j00 vH I~ I Mj
v1 " 7v 7 -- V 1 .3- 3M N 4mNmv S n %
13 N N - H AI HO 0 u I Io- . 1 : U 04 11 133 1H- 1344 1.3 aa9
tp1 41441 0 4 4 N. V34 40 HHH4 av vvvv j wwwH3 o 333 a1 40v4 4 0 11N HHM cON V+000v ) V )ot 4 owm4
w- : 00 u N- 33 cC0 . u uuu u j4)4 .
m vo wr m a o -N mc or 4N m 1 w .mM o0 c0C4 N N lwoo o
41~ ~~~~~~ ooooooo0ooooINN1 3-3. I
33 -4-4.4 I 4 if 341 c3 41 HA-89- -
0.0
'C N lii
" (n 01 94x
14 0000000e
U~~o 0 '.oo'
00o~o 000
.4 1 0 0 0
10 *14 0 00 0
0( H0 0 00 0
.4 a
14 14
(40
22 00;O
9493 4000000
0 (h W9 0000 010Q40 0N0 0 0
0000000U U o occoq40i4
4,:. 0 0: 0 0 0 0 0 0 0 1040 t C44 N4 00 0000f93.
04 a: ooo o
.3 00000C
II~~ 0 4
CN 4 v mN 1
.0~~~2 00p pp-40 Z
44 v 0 0 II (4 (4 w 4N H
04( 0 z11(( 04 0~~~~~ 00 N 04444(( 0 4 14'.4~~~~~ o 0044111.(( 4'H '
41 00 0 14 0. 4 H 14
VC 00000 F4 94 X(v0V 0000000 0a :t4 0z
S40 00000000 0 --
44 4 N C.4Nf44 C~4 0142 N0
H 00 0 0 4 4 ZH(4IW v 'Z A4 00 0 0 9.3 W 4 0 x .0
1t4m 0 00 0 0 (4 (1 1% 0 H00 00 0 0 =01% z (0. U0 000 00 14 D 0.~ V) 0:'C~ ~ 000000 10 0-1(4
94 .4 ( 4 ( wo.44( H
V)4mN14 I3V0 0 4 OU .rt.- 4 '0 40A%1%CM Q I4 H .4-14(4(4(4 1 4 (C
0 V0 g 040( 4 * ~ 00 40 0 H04(( (4 HO a WU.-
a1>4 0 ( QH V >4 (4t .7h (4 I- C4 0 uZ 'Cv t. (4 ---. 404 0 4ID H
CO E(4zN-C -4 .3 m H "I F
'0 14 MN,- Z OX4 4 C0 H H tn--4144* . (4C00 H 1.000o-HO N 4alo 2, H 0 MIII X4 0 04 H0.-4qN 1 i
0 v4v .3 04 0 :11 14 4 ( N ( 0 H N - 14V 9.4
-~~ ~ 0 c(..C . NV114 44 c4 f 0 04 40 1 t)200 0 0 0 o zZ 30D9 uI
1% a 4 u V 0 w L 1. (4444(((( 00 00 -C XL0 140o".-20 aH CE. .A1W04 E 0 4 ,0 V 0 0 o o Q 14 H4 1:. Z. 0(04(Z(H0E.O 14 (4'0
to .4 1 4 0 0 Hco o - CMV 0 0 z r- 14 a U)0 .14 a,0100000 143 0 4~'~'o 04 0 01(0 140H 0 H .14 09
0 0 0: z4 ow 0, -1 00 00 0 :N '.. N- (440( 1 (1z 0-. z C 3--C 1 (. .4 H 14 04 00 0 0 00 4 Nt4-C ( N4~4 H N04 0 00 4 u40 ( 4 0 0 0 0 C ( 0 t0 H .. ~' 4 *4t
14 (4 N 00 4 0. '0 0 4 '0 0000 0 0 ( H1 44 4 (H 4 ' 14 N
Appendix B Sample Output from the Tape Reading Algorithm
B-1
Time Longitude Latitude Brightness Temperature (K)
(Seconds) (Degrees) (Degrees) 19.35v 19.35h 22.235 37v 37b
86543433 55.68 38.46 270.60 266.80 270.40 269.50 265.9086543433 55.81 38.79 272.40 269.30 271.40 269.70 266.7086543433 55.93 39.13 272.10 269.30 271.20 270.50 267.6086543437 55.02 37.48 271.10 268.80 271.60 269.40 266.5086543437 55.17 37.78 270.80 268.20 270.80 269.30 266.1086543437 55.31 38.08 270.00 265.1"0 271.40 268.40 264.3086543437 55.46 38.39 271.10 264.50 269.30 267.40 263.3086543437 55.59 38.72 271.60 267.50 270.10 269.60 266.6086543437 55.71 39.06 271.50 268.60 272.00 269.30 266.0086543440 54.80 37.42 271.90 268.70 270.60 268.60 265.4086543440 54.95 37.72 271.90 268.60 271.60 269.30 266.2086543440 55.09 38.02 271.50 267.60 269.70 268.30 266.0086543440 55.24 38.33 271.00 264.90 269.10 267.60 264.0086543440 55.37 38.66 270.50 265.60 270.20 267.90 263.4086543440 55.49 39.00 . 271.90 268:60 270.30 269.10 266.0086543444 54.58 37.36 271.60 268.20 269.10 268.10 264.9086543444 54.73 37.66 271.20 268.20 269.90 268.50 264.7086543444 54.87 37.96 271.20 267.70 269.90 268.50 266.1086543444 55.02 38.26 271.20 265.90 268.10 268.10 264.9086543444 55.15 38.59 270.60 265.20 269.60 267.80 262.9086543444 55.27 38.93 270.90 266.80 270..00 268.80 264.9086543448 54.35 37.30 270.30 266.70 268.70 267.60 264.8086543448 54.50 37.60 271.10 266.90 267.60 267.90 266.0086543448 54.65 37.90 271.40 267.80 270.40 268.80 264.8086543448 54.80 38.20 270.30 267.30 269.90 267.60 264.4086543448 54.93 38.53 270.00 264.90 268.10 267.10 263.3086543448 55.05 38.86 270.40 264.70 269.80 266.90 262.8086543448 55.18 39.19 270.90 266.00 269.30 268.30 265.1086543452 54.28 37.54 269.80 266.70 268.50 267.20 263.7086543452 54,43 37.83 270.40 266.20 268.30 268.10 264.9086543452 54.58 38.13 270.40 267.20 270.40 268.40 264.9086543452 54.71 38.46 269.70 264.90 268.30 267.20 262.7086543452 54.8-3 38.79 268.80 263.60 268.70 266.90 262.0086543452 54.96 39.12 269.70 263.40 268.50 266.90 261.7086543456 54.21 37.77 270.10 266.60 268.40 266.40 262.8086543456 54.36 38.07 270.30 266.30 269.90 267.30 264.4086543456 54.49 38.39 270.20 265.70 270.10 266.90 264.4086543456 54.61 38.72 269.10 264.30 268.20 265.60 262.1086543456 54.74 39.05 269.20 263.80 269.00 266.60 261.8086543459 54.27 38.33 270.20 265.70 268.80 267.10 263.0086543459 54.39 38.66 269.30 264.00 268.40 266.80 263.5086543459 54.52 38.98 269.00 263.70 268.00 266.40 262.5086543463 54.17 38.59 269.60 264.50 267.40 266.50 262.0086543463 54.30 38.91 269.70 264.30 267.60 266.30 262.5086543467 54.20 39.17 269.80 266.00 268.10 267.30 264.2086720726 57.73 39.10 267.70 264.20 264.90 264.20 262.8086720730 57.70 38.66 266.20 263.70 267.30 265.60 264.3086720730 57.52 38.93 267.50 265.20 267.50 265.40 263.2086720730 57.34 39.18 267.20 264.50 266.80 263.70 261.3086720733 57.68 38.22 264.70 259.80 264.00 263.80 260.5086720733 57.50 38.48 267.60 264.50 267.30 266.40 265.2086720733 57.32 38.75 268.30 265.50 267.30 265.50 264.0086720733 57.14 39.00 267.80 264.90 266.00 265.10 263.2086720737 57.63 37.74 263.50 250.70 264.50 261.80 259.7086720737 57.47 38.04 263.50 258.00 264.90 262.50 257.2086720737 57.29 38.30 267.50 264.60 265.70 266.80 264.9086720737 57.11 38.57 268.30 265.20 267.40 26b.00 265.1086720737 56.93 38.82 267.10 263.70 266.50 264,30 261.20
B-2
Appendix C Mapping Software
c-1
0
.5-H
C400 1
4a>
.4.4 (
010
HIIn
'4. 4 0
I" a4 .4 ., *4 0a A -*
004 * V
.40 tyUI' * -
00 0H34 *r 00
040 V~~ V 54 .5 0if. 40 04 0 V) :1.
Uo Im E. w a,1 x4
VU . v~'5 . Ai I
II~ ~~~ M4 '.-4 *0o4 v ao t. ..4 0 ox
4 045 v* 0 . 1U. * :1 0 w' o-
4 1 Cl 4) '4' 4D iV43 .5 0 4 11 *A 1
IV, 44J * 14 .0Lo 0 V: U. C54UCC:~sJ. W 0 ' 4 913 -'3
.01 0 v 4 'U 0 US a 5 t-4*.5w-40 * .w5 0 a
*~, ..'r. .; -,. o5 0 :0 '5*.4
4 . 0.4 1 43 5 1 " 0 M ) E4.4f4tf! 414- *r w 4 4(4 x" &3 0LaJI5~
V0 O." 4 9 .4 v 0 4t*ty 0 0. 'n I-'U 4 .4 .
*" -34 0 1 :s a- ,1 'I *I 0.aM-4 5 1 ' CN 545'.4 U".4.044~o : 14 H' 0' ' .5C'45-
40. 14 40 06 10 '.4 ..q - 0 0 Dc 3 '-A m II4 ',.WO *%4 'Or0 *1 o c .4 .' . 4 S' r4 0 -4,.' 0 t0@ 43VU
*l 0.410 A , V 5 0 1. 1 0 '54 w40 0I u74
454~1 054 . 5. 1
*N 'U0.SC w 00 H-~ CO mm ow W o 4 w- 0 44o .0 m 0-10 .5; m-ow -*.505~~~~0 -a w4 wN w wo w4 r3 r- U 'U5.aa~~~~~~~~~ .5 . ' f'U u4 U .0.4 - 0I.'40-.4 -
.43444 "4 H .4 ' '5 . *o a.C-2 4
040
00
410
0. ''.
"5(4z
0.:3 e t00 :f.OH4E E , C .4 9 u (
to t ( u a 1 ... A g
tnV t 00C 0wV 7 .X0 i ) 00
Sn -.
00-
0.'
0 o 4 94
0v 0. 00 .0 0 '
00
0. 400004
0 9404900000
C, 000000
a~* 000000>
.40HNNN
:4944 %00 x.
050
401 .3 0
00 E40 H440
%4 4015
A 0 0 0>
00 00 140
-4 to 14 94 a 0-11N£££ 0 000 0
00 oX 0 0000000 X aa44£ ' 000oo00000 000 Lto0 0000000
IIi0 0000000
NN NM
005 ~ ~ - 4' 2 iV
-. >; 0 05 '
> ACX .9- == m4
£0 r 4 V0 0£4V u 4£4z :cUn 4
-14 14*£ )54
0 - 0 M 0O0 Is -3E-~ t0 v. UUU (a~H
940 0.00 40 000400u 00 W04. 0 00000
2.. 0.4v> 0 w0N o 00 0
4, 41 If 0~0 0 41 V o A 0 00000
M 100 '.4JU 0 V o 0 0000000C4~~' v VOV 0 t,0E
Z N '0 0 5.- '0 0000000 o
5.50--
C
0
1.3 9.
.r
4H
H~ U
w1.H 4 n 4 g
x4. t 4 00 k.
z 013 I3x
tn Uf
000 33
0 .0(a W
4*3.414
N ~N
to 14. 0(
0 f 0 z .
>0 .3zN v4 ,
W4 v : P1.m 3
0 )' , i.3 'n H a:
.4 04
U. 4,z 4 .
P.., 0 Na.
t41 0. .3 a HH4 4: . 4 01.3Z H0a 0 3H .
HU9.1 H 3 3 .* 0 ~ 3H 4 H 0u w 0
a 940 0~ 0(30 34 Ia w 0 03 0. n 00, 0 43 a,
o 04NHZOU I' 3 4D r00 9 0 0 -H0 .4 1
W % (av 0141, 014> 4 t- 0 3
o 0: 4 0 H 40 o 4 1.3 0 H
* %W ( - - Z0 0. 4 Mca0 349 a 0.13 I w U Cz 4*3 1.3 0 0-ZO 00 0.3 0W P
A. O 4H t- n0 (4 0~ 04 - 3 Z3 A H m
N w 4F > 3. o.3 1.341.4(3 3m 0 ..Hv o o 0 0 0Z H -03jtr V) a o EH
* 01 0* 03 V H(33~ * 00 H :Z u fo O01 0 06E 0 ; 0K a 1.3 000 0 E0 Y . 0 %4 4~H C4 00 D I -0'::Z.
00 H H 04i 1.3 -A .. U.0H 0Ho 0 a C 1. D- 3 4H 0 H 0. 4 z"F.c
0 'o 1 4 0U.4%30 Qw .1...
H co44 . 3 0 .4 0r.
4 0 04~ 3U4 3'3f
0 ~ HO00H4~Z 0 .
0
mU
00
H.1.0aa
> o(Y0 c 1
C4 : 44 .10100
4v 0 -
0co
x x 01 000
I(h 00-0 .0 000
z z 0 00000~C 0 00 C400
040 UUU . 000
040) 0 0 -A 0 003r! 1, 94 0 0 0 1
0 . . I000x~ 01100 410 : zzC
.0 0 0 00>I.4 v~ v 0
41 x w L.
:4If C60
*0 0.u
-0 LAC M
0 4) 0 ,
* -- 0. w~ 42-
041 (.0* 100o , ..4 -A4 0 .
D1 a4 -4 0 -4 O C
0 U) 0 0 000 0-IU 0- U. OC. 0
4. , .O fa C .
to 0 .C, *U' .4 0 0-u0 C4.* 4, o Ci 00 44 xHH
0i X. * . 4 V. * .u -5 3 01x01:3 000 z*I .3 x
:1 0 u t4l* C aC .
*.4 0 *(I O O.4 * .
0- 0 E-. (i
a:J~ Z vs0 0 0.0
.. Io 0.O U) 44 - 4 C-i H .4t ,Ii 00 -I0 0 w.4 11 00.4, 000 a'0 4 V o 0'C00
:4 *...1 v. C a. 0 0m4a 0
0 0 0 w44 *0 4 N v0v-IC0 0 0 I 1*0 04 11 MM4.
I 0 00 0 000000 >U ' O S 00
0 Ii 0~ '4C0 C C-6
0
0
(4
0.=In 3H
0ac4 L
! f
In
o n
104
010 0
01 C4 .3
z02Ne4.f w m
v~ to
0-0 .4.-I >014z
E.z
0: 0 .3
142i2Z 14
z 0 0 z 1 p: 4 H )co 0 010 00
0 :10 t4n a, H " 014U3 znC - 14i1* 0
0 w V) 10. 140uA=0 z O 940
0 0 049O.H .x0 .3m oI.. 01 H 4 0 " WI .
z .3 A 14InZ f 0 0
>. ogpg.4 z H
0 "; C.. z 44 A.. N E
o AZ W9(4 X H1~a1(4 U 1 14Ck.OC4 01.1
00 0 C3Z 0140Z1 - 00 H t44I, to iz14 V a 0 0 :. Zin 0 L H 1 4 .O 0
E!. Z 1 14-004 00's 0W1 00 U'4 % - - f ! oa Dz n 0 4140 0
z1 N CiUiin z~ 2 C.o w .01 ou4z 1 1.."OH 4 t" u-4 .-
0n 0 In .. 2."0H010 aH COP"Q. 0. H W000-HON r. *0000 1z ww3 a 0141110 . z a. 4CIC41
N, 14 z C4"H~ 94IW 001 011~01 1
Ov v . 00.a4 . 1
01 I In 4 In4Z01-4E4 0-7 I
0
0
M 9.
0.
00
0.0H
00
50z H
01
0~0 0 V.
.4 000 0 0
cic000 'a 0f0Hzza .0 Coco
O '4000
C:~~0 * 94 *:7 1 *
10 '3 000 N. N m~MO~t autn0 W w
mw 00z
0- 'a 0000 c
UU'o 0 :3 liltz'a Si 41 0 00
II 0C 1) 01 (-4 %4 i 0 H Z
41 .14.40. 40 m4 Vuu
ds . 0 1.
jo . 0. ta m .4.m 400'4) 0 ; 41 C4E.1
0 00 Si 0
O3g: 1 41' -. 4
.4 SCO'44 . 00 4 ~si 004 0 . uA i 0 a ' U U
.40 ".'4 i 04' H01 Ai L-4 1% r Z..
0 '0 0.4'
ID 00 m. V.00o i ul 0 044.04
u -ox P. -0 fl Fi t-'0 H D4. ' 0 4 u 0 ,1' 0'0
:3.4 N tl04 4' .4 Z4 z '0Scoj.3t
o r Sz .0 4 'u,0 :20 ' 0i t) I * 4'
a -. i . *0 1:,ow(, t C
*~~C 54 5' * *.
z' aa 05S0 a, *4 4 a 044'a 00044 '4 0.0 40. 0US -4.00 0 0
9 Si 00 V0.- *M c 000
C4 " N. . 0 *0 "Mm"
00000i 00 0 0 00 0 00000 C61. w >0124.
I a' ~ .0-- *8
0
0.
0.
1..4
a:0
0.
00m0%.
101
'a':2 H
t41. Z -1140
z1 V0 4OHl 00.
U-4 00 0
40 0 U4 0
0- 0 r m.,t
9' coDoVZ O0
0.? 00 0:w l
0)00o o
14.. 01 J -. 4
V. >014=4 zo: 01 -4 Z .444
tj Z ZHOO040
0o 0 ,3o0o mo
0 0
z 4 woo o H4 4m E40 CZZF .4 0 H. [4 C 4V
Z.~ Zm o g r4 m 0N 0. .m 1ww H414H H 9: 4dt
* :. . 0, .4owc fz o H) :30 4 w ~ 01 NI 122 01c a
3. 0I X 0.01HZ. v
0. C-1N
C. 0
0 0 X
.A H'
ta.
.44
in 0
m 04 0
.4~9 wx 0
0C
U~v H
4.,.
0:4 94
0v:0
.E4
Z4' 0 X
4: m4. 4'
:1 0:
v'4 - 14 1'
.40
00 X 0 0U=:3 A. Ca~ 9 0
0. a. 0 0i
vi" uC e )
0 4 E4 C
.4.
L* 0 0 0 0 M
.a a. > vv o
400 0 0 0 " 4 m f.4 1%4I
... . .. ., 000000 0000 0
4 II U10
Appendix D Sanple Mapping Software Output
D-1
Lu --- -- ---- x
I __ It
Alf~~
Iffl >Q~.~t1 *D-2
'4
441
It f
S D" -3-. -I . . ... .. .
4-/-- - / .... - ..._
1x ""-']
* /. I
%-\,.
i
;
I
.
/
4 ",
' N ,
D-3
• i. I
I , I
/
I?',
\ "-://' .
D-4
Appendix E SSMfi Data Catalog
E-1
SSM/I Data Received
Tape #t Beginning of dataset End of dataset # files
01 01-jun--89:00:07:31 05-jun-89:00:57:56 5602 05-jun-89:00:57:56 09-jun-89:00:06:24 2003 09-jun-89:00:06:24 13-Jun-89:00:56:52 5504 13-Jun-89:00:56:53 16-jun-89:00:18:21 4205 16-jun-89:00:18:21 20-jun-89:01:08:51 4906 20-jun-89:01:08:51 24-jun-89:00:17:17 5407 24-jun-89:00:17:18 28-Jun-89:01:07:33 4908 28-jun-89:01:07:33 01-jul--89:00:28:4 4 4209 01-jul-89:00:28:55 05-jul-89:01:18:59 4910 05-jul-89:0l:18:58 09-Jul-89:00:27:20 5311 09-Jul-89:00:27:20 13-Jul-89:01:17:35 5412 13-Jul-89:0l:17:35 16-jul-89:00:38:48 4213 16-jul-89:00:38:40 20-jul-89:01:29:01 5714 20-Jxul-89:01:29:01 24-jul-89:00:37:17 1215 24-jul-89:00:37:17 28-jul-89:01:27:89 3216 28-Jul-89:01:27:29 01-Aug-89:00:35:45 5617 01-Aug-89:00:35:45 05-Aug-89:01:25:54 5718 05-AUg-89:01:25:54 09-Aug-89:00:34:07 5119 09-Aug-89:00:34:07 13-AUg-89:01:24:18 5720 13-AUg-89:01:24:19 16-Aug-89:00:45:28 4221 16-Aug-89:00:45:28 20-Aug-89:01:35:36 5622 20-Aug-89:01:35:36 24-Aug-89:00:43:47 5323 24-AUg-89:00:43:47 28-Aug-89:01:33:51 5724 28-Aug--89:01:33:51 01-Sep-89:00:41:48 5625 0l-Sep-89:00:41:46 05-Sep-89:01:31:45 5526 05-Sep-89:01:31:45 09-Sep-89:00:39:47 569 7 09-Sep-89:00:39:47 13-Sep-89:01:29:43 5728 13-Sep-89:01:29:43 16-Sep-89:00:50:41 4229 16-Sep-89:00:50:41 20-Sep-89:01:40:33 5730 20-Sep-89:01:40:33 24-Sep-89:00:48:26 5631 24-Sep-89:00:48:26 28-Sep-89:01:38:15 5732 28-Sep-89:01:38:15 01-Oct-89:00:59:09 40
There is no data missing from our requested set.
E-2